feat(work): add Set Matrix Zeroes solution

feat(api): add /admin/solved endpoint and commit source handling

.claude/skills/campaign-issues/SKILL.md

@@ -0,0 +1,175 @@
+---
+name: campaign-issues
+description: Create GitHub issues for this LeetCode campaign — individual problem issues and the topic issues that own them — using the repo's exact title, label, milestone, body, and sub-issue conventions. Use when asked to add/create an issue for a LeetCode problem or a topic, attach a problem to its topic, or backfill the issue catalog.
+---
+
+# Campaign issues
+
+The GitHub issue tracker **is the campaign catalog**: ~24 `topic` issues each own a
+set of `problem` sub-issues. The Worker reconciles this catalog into D1 nightly
+(`/admin/reconcile`), so a correctly-labelled issue is all that's needed — the SRS,
+digest, charts, and Project mirror pick it up automatically. Get the labels/body shape
+right; everything downstream is reconciliation.
+
+All commands use `gh` against `prdlk/leetcode` and fall back to `gh auth token`.
+
+## Problem issue
+
+One issue per LeetCode problem, attached as a sub-issue of its topic.
+
+- **Title:** `LC <num> · <Title>` — the LC number and the problem's name, nothing
+  else. Separator is a middle dot `·` (U+00B7) with single spaces. Difficulty and
+  set are **labels only**; never repeat them in the title (they used to be a `·
+  Hard · core` suffix — that is gone, and `catalog.ts` no longer parses it).
+- **Labels** (four, all required — these are what the Worker reads):
+  - `problem` — marks the sub-issue as curriculum; without it the reconcile skips it.
+  - `diff:easy` | `diff:medium` | `diff:hard` — LeetCode's own rating, do not invent.
+  - `set:core` | `set:optional` | `set:deferred` — see set semantics below.
+  - `phase:N` — the **topic's** phase (same phase label the topic issue carries).
+- **Milestone:**
+  - `core` / `optional` → the topic's phase milestone.
+  - `deferred` → `Review — Transfer & Mocks` (deferred Hards are parked until week 8).
+  - Pass the milestone **title**, not its number, to `--milestone`.
+- **Body** (exact shape; `<slug>` is the LeetCode URL slug, `<topic-issue>` the owning topic's number):
+
+  ```
+  https://leetcode.com/problems/<slug>/
+
+  - **Topic:** #<topic-issue>
+  - **Set:** <core|optional|deferred>
+  - **Language:** Python
+
+  ### Close-out (paste before closing this issue)
+  - Time: O(?)
+  - Space: O(?)
+  - Edge case traced:
+  - First attempt passed: yes / no
+  - Miss note:
+  ```
+
+Create it:
+
+```sh
+gh issue create --repo prdlk/leetcode \
+  --title "LC 42 · Trapping Rain Water" \
+  --label problem --label set:core --label diff:hard --label phase:1 \
+  --milestone "Phase I — Linear Structures" \
+  --body "https://leetcode.com/problems/trapping-rain-water/
+
+- **Topic:** #9
+- **Set:** core
+- **Language:** Python
+
+### Close-out (paste before closing this issue)
+- Time: O(?)
+- Space: O(?)
+- Edge case traced:
+- First attempt passed: yes / no
+- Miss note:"
+```
+
+Then **attach it as a sub-issue of the topic** (this is a separate API call — creating
+the issue does not link it). The sub-issues API wants the child's internal **id**, not
+its issue number, and the field is integer-typed, so use `-F` (not `-f`):
+
+```sh
+child_id=$(gh api repos/prdlk/leetcode/issues/211 --jq '.id')
+gh api --method POST repos/prdlk/leetcode/issues/9/sub_issues -F sub_issue_id=$child_id
+# verify
+gh api repos/prdlk/leetcode/issues/9/sub_issues --jq '.[] | "\(.number)  \(.title)"'
+```
+
+## Topic issue
+
+One issue per curriculum topic; owns the problem sub-issues.
+
+- **Title:** `Topic NN — <Name>` (zero-padded number, em dash, e.g. `Topic 07 — Monotonic Stack`).
+- **Labels:** `topic`, `phase:N`.
+- **Milestone:** that phase's milestone (title, not number).
+- **Body:**
+
+  ```
+  **Scheduled:** <Weekday Mon DD, YYYY>
+  **Phase:** <Roman> — <Phase name> (Week N)
+  **Builds on:** Topic <n>
+
+  ### Trigger
+  TODO
+
+  ### Invariant
+  TODO
+
+  ### Trap
+  TODO
+
+  ### Close-out ritual
+  Before every submit, out loud:
+  1. State time and space complexity.
+  2. Trace one edge case.
+
+  ### Day rules
+  - 90-minute coding cap. Stop at the cap, even mid-problem.
+  - Core problems first. Optional only if the core is done.
+  - Finished early? Do this topic's optional set. Do not start tomorrow.
+  ```
+
+`Scheduled` should come from the calendar of record — `apps/api/data/schedule.json`
+(day → topic issue) — not guessed.
+
+## Reference tables
+
+Set semantics (from the label descriptions):
+
+| set | meaning | milestone |
+|---|---|---|
+| `core` | must solve on the topic day | topic's phase milestone |
+| `optional` | solve only if the clock allows | topic's phase milestone |
+| `deferred` | a Hard, parked until week 8 | `Review — Transfer & Mocks` |
+
+Milestones (`--milestone` takes the **title**):
+
+| # | title |
+|---|---|
+| 1 | Phase I — Linear Structures |
+| 2 | Phase II — Nodal & Grid |
+| 3 | Phase III — Hierarchical |
+| 4 | Phase IV — Relational |
+| 5 | Phase V — Decision Space |
+| 6 | Review — Transfer & Mocks |
+
+Find a topic issue and confirm its phase before creating problems under it:
+
+```sh
+gh issue list --repo prdlk/leetcode --label topic --state all --limit 30 \
+  --json number,title,labels,milestone
+```
+
+## Gotchas
+
+- **Check the LC number is not already in the curriculum BEFORE creating.** D1 keys
+  `problems` on `lc_number` (`ON CONFLICT(lc_number)`), so two issues for one LC number
+  silently fight over a single row — whichever the reconcile walks last wins, and the
+  loser's issue number ends up orphaned in D1. A problem can legitimately move topic or
+  set; that is a relabel/re-parent of the *existing* issue, never a second issue:
+
+  ```sh
+  gh issue list --repo prdlk/leetcode --label problem --state all --limit 300 \
+    --search "LC 42" --json number,title,labels,state
+  ```
+
+  If a row already exists, change the existing issue instead: swap its `set:*` label,
+  move the sub-issue to the new topic, and fix its milestone. Then reconcile the board
+  (`project-fields` skill). If a duplicate did get created and one is deleted, the
+  surviving issue number and set only reach D1 on the next catalog reconcile — SRS
+  columns (`stage`, `next_review`) are preserved, since they are the real attempt history.
+
+- **`-F` vs `-f`:** `sub_issue_id` must be an integer; `-f` sends a string and the API
+  rejects it with `is not of type integer`.
+- **id, not number:** the sub-issue payload wants the child's `.id` (a large opaque
+  integer), never its display `#number`.
+- **`--milestone` wants the title string**, not the numeric id (`'1' not found` otherwise).
+- **Reconcile, don't react:** don't edit D1 or the Project by hand — a correct issue is
+  reconciled into SRS state on the nightly cron. Solutions landing in `work/` are what
+  close problem issues (via `close-solved.yml`), which then roll up to close topics.
+- **Difficulty is LeetCode's**, `set` is our scheduling choice — they're independent. A
+  Hard can be `core` if we deliberately schedule it (vs the usual `deferred`).

.claude/skills/project-fields/SKILL.md

@@ -0,0 +1,108 @@
+---
+name: project-fields
+description: Reconcile the "Interview Prep" GitHub Project (user project #2) fields — Set and Difficulty projected from issue labels — and understand which fields the Worker mirrors vs which live only in D1. Use when asked to sync/fix/backfill Project fields, when Project columns look empty or stale, or when adding a field to the project.
+---
+
+# Interview Prep Project fields
+
+The user-level Project **"Interview Prep"** (`prdlk` project #2, id `PVT_kwHOADoNjs4BhV1a`)
+is a *board over* the issue catalog, never a source of truth. Every field is a projection
+of something owned elsewhere.
+
+## Field ownership — the rule that matters
+
+| Field | Owner | Written by |
+|---|---|---|
+| `Set` (Core/Optional/Deferred) | issue label `set:*` | `sync-project-fields`, last step of `close-solved.yml` |
+| `Difficulty` (Easy/Medium/Hard) | issue label `diff:*` | `sync-project-fields`, last step of `close-solved.yml` |
+| `Target Date` | D1 `next_review` | Worker mirror, per log (`apps/api/src/mirror.ts`) |
+| `Status`, `Labels`, `Milestone`, `Parent issue`, … | GitHub built-ins | GitHub |
+
+**SRS stage and first-attempt result are deliberately NOT Project fields.** They live only
+in D1; the charts, digest and `/api/stats` read them from there. Duplicating them into
+Project single-selects bought nothing but drift and two extra writes per log. Do not re-add
+them to the project, and do not re-add `setStage`/`setFirstAttempt` to `mirror.ts`.
+
+## Reconciling Set / Difficulty
+
+`apps/cli/sync-project-fields.ts` recomputes every desired value from the issue labels and
+writes only what differs — same reconcile-don't-react shape as `close-solved`.
+
+**In CI it runs itself.** It is the final step of `.github/workflows/close-solved.yml`, so
+any push under `work/` reconciles the board along with the issues and D1. The step is
+`continue-on-error: true` and skips cleanly when `PROJECT_PAT` is unset — D1 and the issues
+are the truth, so a Project failure must never fail a run that already closed issues and
+logged the solve. It honours the workflow's `dry_run` input via `DRY_RUN=true`.
+
+Run it by hand when issues changed without a `work/` push (e.g. you just filed a problem
+issue, or relabelled one):
+
+```sh
+# needs a token with the `project` scope — neither `gh auth token` nor GITHUB_TOKEN has it.
+export GH_TOKEN=$(grep '^GH_PAT=' apps/api/.dev.vars | cut -d= -f2-)
+
+bun run sync-project-fields -- --dry-run   # or DRY_RUN=1|true; prints the table, writes nothing
+bun run sync-project-fields                # apply
+bun run sync-project-fields                # re-run: must report 0 writes
+```
+
+Invariants the script holds — preserve them in any edit:
+
+- **Only `problem`-labelled items are touched.** Topic, cadence and review issues have no
+  set/difficulty; they are skipped, never written and never cleared.
+- **Idempotent.** A second run reports `0 field value(s)`. That re-run *is* the verification.
+- **Labels are truth.** Never "fix" a field by hand in the UI — fix the label, then reconcile.
+  A hand-edited field is silently reverted on the next run.
+- **Fields and options resolve by NAME**, so re-creating a field in the UI doesn't strand a
+  dead id. A missing field or option is a hard error, not a silent skip.
+- **A problem missing a `set:`/`diff:` label** warns and is left untouched (it means the
+  issue is mislabelled — fix it with the `campaign-issues` skill).
+
+## Auth gotcha
+
+Project v2 needs the `project` scope. The local `gh` login has only
+`gist, read:org, repo, workflow`, so **any** `gh project`/GraphQL project call fails with
+the default token. Use the Worker's `GH_PAT` from `apps/api/.dev.vars` via `GH_TOKEN`
+(never echo it). Symptom of the wrong token: `Resource not accessible by personal access
+token` or an empty `projectsV2.nodes`.
+
+## Inspecting the project
+
+```sh
+export GH_TOKEN=$(grep '^GH_PAT=' apps/api/.dev.vars | cut -d= -f2-)
+
+# fields + single-select options
+gh api graphql -f owner=prdlk -f title="Interview Prep" -f query='
+query($owner: String!, $title: String!) {
+  user(login: $owner) { projectsV2(first: 10, query: $title) { nodes { id title number
+    items(first: 1) { totalCount }
+    fields(first: 30) { nodes {
+      ... on ProjectV2FieldCommon { id name dataType }
+      ... on ProjectV2SingleSelectField { options { id name } } } } } } }
+}' --jq '.data.user.projectsV2.nodes[] | select(.title=="Interview Prep")
+         | .fields.nodes[] | "\(.name)\t\(.dataType)\t\((.options // []) | map(.name) | join(","))"'
+```
+
+Writes use `updateProjectV2ItemFieldValue` with
+`value: { singleSelectOptionId: … }` (dates use `{ date: "YYYY-MM-DD" }`); clearing uses
+`clearProjectV2ItemFieldValue`. An item id is *not* an issue number — resolve it from the
+project's `items` connection (or the issue's `projectItems`), as `mirror.ts` does.
+
+## Adding a new projected field
+
+1. Create the field + options in the Project UI.
+2. Add it to the `FIELDS` table in `apps/cli/sync-project-fields.ts` — one entry mapping
+   label → option name. Nothing else needs to change; resolution, diffing, dry-run and
+   reporting are generic over that table.
+3. Dry-run, apply, re-run to prove idempotency.
+
+If the new field's truth lives in **D1** rather than a label, it belongs in the Worker
+mirror instead (`apps/api/src/mirror.ts`) — but read the ownership rule above first and
+default to *not* mirroring.
+
+## Known trap
+
+`resolve()` in `mirror.ts` throws if a field it asks for is absent, *before* caching
+`info` — so one missing field breaks **every** mirrored write, including `Target Date`,
+and only as a `console.warn`. If Target Date silently stops updating, check that every
+field `resolve()` looks up still exists in the project.

.github/workflows/close-solved.yml

@@ -28,7 +28,36 @@ jobs:
       - uses: oven-sh/setup-bun@v2
 
       # No install step: the script only uses Bun builtins + fetch.
-      - run: bun scripts/close-solved.ts
+      #
+      # SRS_ADMIN_KEY lets the script report the solved set to the Worker
+      # (POST /admin/solved). Closing an issue is invisible to D1 — the
+      # catalog reconcile never reads issue state, and only logAttempt() moves
+      # the ladder — so without this a pushed solution would never reach the
+      # charts, the digest's solved ticks, or the drill/gate pools.
+      - run: bun apps/cli/close-solved.ts
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
           DRY_RUN: ${{ inputs.dry_run }}
+          SRS_ADMIN_KEY: ${{ secrets.SRS_ADMIN_KEY }}
+
+      # Project fields are a projection of the issue labels, so the same push
+      # that adds a solution (and any issue filed since the last run) can leave
+      # the board's Set/Difficulty empty. Reconciling here keeps the one
+      # "solution landed" entry point responsible for the whole fan-out; the
+      # script is idempotent, so a run with nothing to do writes nothing.
+      #
+      # Needs the `project` scope, which GITHUB_TOKEN does not have — hence
+      # PROJECT_PAT. Unset = skip, matching SRS_ADMIN_KEY above. Best-effort by
+      # design: D1 and the issues are the truth, so a Project failure must not
+      # fail a run that already closed issues and logged the solve.
+      - name: Reconcile Project Set/Difficulty
+        continue-on-error: true
+        env:
+          GH_TOKEN: ${{ secrets.PROJECT_PAT }}
+          DRY_RUN: ${{ inputs.dry_run }}
+        run: |
+          if [ -z "$GH_TOKEN" ]; then
+            echo "PROJECT_PAT unset — skipping the Project field reconcile"
+            exit 0
+          fi
+          bun apps/cli/sync-project-fields.ts

.github/workflows/close-topics.yml

@@ -0,0 +1,49 @@
+name: Close Topics
+
+# No D1 push here, unlike Close Solved: D1 stores no topic completion column.
+# Phase progress, the gate pools and the charts all derive from the problem
+# rows, which close-solved already logs — a topic is finished exactly when its
+# core problems are, so there is nothing left to reconcile.
+
+on:
+  # Close Solved closes problem issues with GITHUB_TOKEN, and token-driven
+  # closes never fire `issues` events — so chain off its run instead.
+  # Conclusion is not filtered: the script reconciles state, so running after a
+  # partially successful run is safe and catches whatever it did close.
+  workflow_run:
+    workflows: [Close Solved]
+    types: [completed]
+  # Covers problem issues closed by hand in the UI.
+  issues:
+    types: [closed]
+  workflow_dispatch:
+    inputs:
+      dry_run:
+        description: Report finished topics without closing anything
+        type: boolean
+        default: false
+
+permissions:
+  contents: read
+  issues: write
+
+concurrency:
+  group: close-topics
+  cancel-in-progress: false
+
+jobs:
+  close:
+    # The `issues` trigger fires for every issue in the repo; only closing a
+    # problem sub-issue can ever complete a topic's core set.
+    if: github.event_name != 'issues' || contains(github.event.issue.labels.*.name, 'problem')
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: oven-sh/setup-bun@v2
+
+      # No install step: the script only uses Bun builtins + fetch.
+      - run: bun apps/cli/close-topics.ts
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          DRY_RUN: ${{ inputs.dry_run }}

.github/workflows/deploy.yml

@@ -1,4 +1,4 @@
-name: Deploy Docs
+name: Deploy
 
 on:
   push:
@@ -10,8 +10,11 @@ permissions:
   pages: write
   id-token: write
 
+# The Worker deploy is idempotent (wrangler uploads the whole bundle), so on a
+# burst of pushes only the last one matters: cancel-in-progress is correct for
+# both the Pages build and the Worker job.
 concurrency:
-  group: pages
+  group: deploy
   cancel-in-progress: true
 
 jobs:
@@ -31,7 +34,7 @@ jobs:
 
       - uses: actions/upload-pages-artifact@v3
         with:
-          path: dist
+          path: apps/docs/dist
 
   deploy:
     needs: build
@@ -42,3 +45,24 @@ jobs:
     steps:
       - id: deployment
         uses: actions/deploy-pages@v4
+
+  # Deploys on every main push (not path-filtered): apps/api/data/schedule.json
+  # is bundled at deploy time, and re-deploying an unchanged Worker is a no-op.
+  # Needs repo secrets CLOUDFLARE_API_TOKEN (Workers Scripts: Edit) and
+  # CLOUDFLARE_ACCOUNT_ID.
+  deploy-api:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: oven-sh/setup-bun@v2
+
+      - run: bun install --frozen-lockfile
+
+      # DST cron guards must hold before the Worker ships.
+      - run: bun run api:test
+
+      - run: bun run api:deploy
+        env:
+          CLOUDFLARE_API_TOKEN: ${{ secrets.CLOUDFLARE_API_TOKEN }}
+          CLOUDFLARE_ACCOUNT_ID: ${{ secrets.CLOUDFLARE_ACCOUNT_ID }}

.github/workflows/spaced-repetition.yml

@@ -0,0 +1,54 @@
+name: Spaced Repetition
+
+# Opens (or refreshes) today's `Spaced Repetition — <Month D, YYYY>` issue from
+# whatever the SRS ladder has due (`POST /admin/due`, the same rows the 8 AM
+# digest mails — so the two cannot disagree). The script stores nothing and
+# schedules nothing: it asks D1 what is standing today, keyed by the ET date in
+# the title, so a re-run rewrites one body instead of stacking duplicates.
+# Sunday is the rest day and the ladder never books it, so nothing is written.
+
+on:
+  schedule:
+    # 10:00 UTC = 06:00 ET while the campaign runs (EDT through Nov 1), 05:00 ET
+    # after. No DST guard and no second cron: the issue is keyed by ET calendar
+    # date, not by the hour it was created, so the drift is cosmetic.
+    - cron: "0 10 * * *"
+  workflow_dispatch:
+    inputs:
+      dry_run:
+        description: Print the issue body without creating or updating anything
+        type: boolean
+        default: false
+      date:
+        description: ET date to treat as today (YYYY-MM-DD) — backfill a missed day
+        type: string
+        default: ""
+
+permissions:
+  contents: read
+  issues: write
+
+concurrency:
+  group: spaced-repetition
+  cancel-in-progress: false
+
+jobs:
+  drill:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: oven-sh/setup-bun@v2
+
+      # No install step: the script only uses Bun builtins + fetch.
+      #
+      # SRS_ADMIN_KEY reads today's due list from the Worker. Unset is fatal
+      # here, unlike in close-solved: without D1 there is no schedule to render,
+      # and inventing one from issue close dates is exactly the second source of
+      # truth this replaced.
+      - run: bun apps/cli/spaced-repetition.ts
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          DRY_RUN: ${{ inputs.dry_run }}
+          SRS_TODAY: ${{ inputs.date }}
+          SRS_ADMIN_KEY: ${{ secrets.SRS_ADMIN_KEY }}

.github/workflows/sync-d1.yml

@@ -0,0 +1,36 @@
+name: Sync D1 Catalog
+
+# GitHub issues own the SRS catalog (topics, problems, set/diff labels,
+# milestones); the Worker's D1 holds a reconciled copy. The 8 AM ET cron is
+# the backstop — this workflow pushes curriculum edits into D1 immediately.
+#
+# The reconcile is a total recompute (never touches SRS-owned columns), so
+# bursts can safely collapse to the last run: cancel-in-progress is correct.
+
+on:
+  issues:
+    types:
+      [opened, edited, labeled, unlabeled, milestoned, demilestoned, reopened, transferred, deleted]
+  workflow_dispatch:
+
+permissions: {}
+
+concurrency:
+  group: sync-d1
+  cancel-in-progress: true
+
+jobs:
+  reconcile:
+    # Only curriculum issues shape the catalog.
+    if: >-
+      github.event_name == 'workflow_dispatch' ||
+      contains(github.event.issue.labels.*.name, 'problem') ||
+      contains(github.event.issue.labels.*.name, 'topic')
+    runs-on: ubuntu-latest
+    steps:
+      - name: Reconcile catalog into D1
+        run: |
+          curl -sf -X POST -H "Authorization: Bearer $SRS_ADMIN_KEY" \
+            https://srs-api.prdlk.workers.dev/admin/reconcile
+        env:
+          SRS_ADMIN_KEY: ${{ secrets.SRS_ADMIN_KEY }}

.gitignore

@@ -1,6 +1,8 @@
 # dependencies (bun install)
 node_modules
 tmp/
+*.log
+.crush
 
 # output
 out
@@ -35,5 +37,10 @@ report.[0-9]_.[0-9]_.[0-9]_.[0-9]_.json
 .DS_Store
 /CLAUDE.md
 .blume/
-/docs/.wrangler/state/v3
+/apps/docs/docs/.wrangler/state/v3
 .blume-verify/
+
+# api worker
+apps/api/.dev.vars
+apps/api/.wrangler/
+apps/api/worker-configuration.d.ts

.omp/commands/sync-solutions.md

@@ -1,35 +0,0 @@
----
-description: Sync work/ leetcode solutions into docs/solutions/ pages (gold standard format)
----
-
-Sync every leetcode solution under `work/` into a docs page under `docs/solutions/`, formatted like the gold standard `docs/solutions/(two-pointers)/344-reverse-string.mdx`. Extra focus (optional): $@
-
-## Run the script
-
-The port is fully scripted — do not hand-write pages:
-
-```
-bun run sync
-```
-
-`scripts/sync.ts` does the whole pipeline:
-
-- Globs `work/**/*.{js,py}` (`<number>.<kebab-slug>.{js,py}` under `work/<Difficulty>/<Category>/`) and groups the two language variants of a problem by leetcode number.
-- Parses each file's header comment (js block comment / py docstring) into title, difficulty, statement, examples, constraints, and follow-up.
-- New problems get a full page at `docs/solutions/(<category>)/<leetcode-number>-<kebab-slug>.mdx`: frontmatter, `<Badge>` topic (README curriculum table, falling back to the `work/` subdirectory), `::::warning` for follow-ups/special requirements, examples, constraints, and the solution code verbatim.
-- `(<category>)` is the kebab-cased `work/` subdirectory (`(array)`, `(hash-table)`, `(two-pointers)`, …) and becomes the sidebar group; the leetcode-number filename prefix orders pages numerically. Both are stripped from the URL (`(two-pointers)/344-reverse-string.mdx` → `/solutions/reverse-string`). The script moves any page whose path drifts from this scheme.
-- Frontmatter `title` keeps the number (`'344. Reverse String'`); `sidebar.label` is the number-free problem name. The script backfills a missing `sidebar.label` on existing pages but never overwrites one.
-- Already-ported pages (frontmatter title starts with the leetcode number) only get their `## Solution` section regenerated — curated prose is never touched.
-- Problems solved in both languages render as a `<CodeGroup>` (Python first, then JavaScript); single-language solutions render as a plain fence.
-
-The script prints the report table: work file → docs page → created / moved / updated / skipped (already in sync).
-
-## Review
-
-Skim any **created** pages — the prose transforms (backticking literals in example explanations, warning detection) are heuristic. Fix wording by editing the page prose directly; it won't be overwritten on the next sync. Never edit inside `## Solution` — that section is owned by the script and synced from `work/`.
-
-## Verify
-
-1. `bunx blume build --isolated` must succeed with no new warnings; confirm each new route generated (`/solutions/<slug>`).
-2. Spot-check one new page's built HTML for the badge, callout (when present), and solution code.
-3. `rm -rf .blume-verify` when done.

AGENTS.md

@@ -0,0 +1,121 @@
+# Repository Guidelines
+
+## Project Overview
+
+LeetCode interview-prep campaign (8 weeks, 2026-08-17 → Oct 11) run as a repo: solution files under `work/`, a Blume docs site under `apps/docs/`, GitHub Actions that reconcile issues and publish docs, and a Cloudflare Worker (`apps/api/`) that runs the spaced-repetition system (SRS) — daily digest email, one-tap logging, weekly review issues, live charts. Issues model the curriculum: ~24 `topic` issues (`Topic NN — Name`) own ~161 `problem` sub-issues titled `LC <num> · <Name>`, with difficulty and set carried **only** by labels (`set:core|optional|deferred`, `diff:*`) — never by the title; milestones are phases; a user-level GitHub Project ("Interview Prep", #2) mirrors review state.
+
+## Architecture & Data Flow
+
+Three owners, one direction of truth:
+
+```mermaid
+flowchart LR
+  work[work/ solutions] -->|bun run sync| docs[apps/docs/content/*.mdx]
+  work -->|close-solved.yml| PI[problem issues]
+  work -->|close-solved.yml POST /admin/solved| D1[(D1 = SRS state)]
+  PI -->|close-topics.yml| TI[topic issues]
+  GH[GitHub issues = catalog] -->|nightly reconcile| D1
+  D1 -->|8 AM ET cron| Mail[digest email + one-tap links]
+  Mail -->|GET /log| D1
+  PI -->|/done webhook| D1
+  D1 -->|best-effort mirror: Target Date only| Proj[GitHub Project fields]
+  GH -->|close-solved.yml sync-project-fields| Proj
+  D1 -->|Sat midnight ET| Review[Review — Week N issue]
+  PI -->|"spaced-repetition.yml (6 AM ET cron, closed +3d/+7d)"| SR[Spaced Repetition — Month D issue]
+  D1 --> Charts[SVG charts + /api/stats]
+  Charts -->|bun run stats| TUI[terminal dashboard]
+```
+
+- **Reconcile, don't react.** `close-solved.ts`, `close-topics.ts`, and the Worker's catalog sync recompute desired state from scratch each run: re-runs are no-ops, backfills need no special casing, closing is one-directional. `close-solved` posts its *whole* implemented set to `/admin/solved`, not just the issues it closed this run.
+- **Solving has three doors, one write path.** A digest one-tap, a `/done` comment, and a solution landing in `work/` all end in `logAttempt()`. Closing an issue is *not* one of them: the catalog reconcile never reads issue state, so a `work/` push that skipped the email would otherwise leave `stage='new'` — invisible to the charts, the digest's solved ticks, and the drill/gate pools. Hence `/admin/solved`, which takes `{lc, bucket}` pairs (`source='commit'`): the bucket names the rung the file settles — `work/1` settles `new`, `work/3` settles `+3`, `work/7` settles `+7` — so a pushed re-solve advances the ladder instead of vanishing, and re-posting the whole implemented set writes nothing because the rung named has already been left behind. That stage check is the entire idempotency mechanism; no index, no dedupe table, no state file. `close-topics` needs no such call — D1 stores no topic completion, only problem rows.
+- **The rest day is structural, not cosmetic.** Sunday is never booked: topics run Mon–Fri (`apps/api/data/schedule.json`), the gate is Saturday, and the ladder's windows are 3 and 7 precisely because those return a solve to a working day. Every scheduled date — ladder review, levelled overflow, deferred-Hard release, review-issue window — comes out of `workingDay()` in `apps/api/src/srs.ts`, which slides the one exception (a Thursday solve's `+3`) forward to Monday. Forward, never back: a review may slip later than its interval, never shorten it. Never mint a scheduled date with bare `addDays()`, and never "fix" a Sunday landing downstream — the digest's Sunday rest branch is a courtesy, not the mechanism. `apps/api/src/srs.test.ts` sweeps the campaign calendar to keep this honest.
+- **One schedule, one answer to "what is open today".** D1's ladder IS the schedule; `dueToday()` (`POST /admin/due`) is its read side and `DUE_WHERE` is the single SQL definition of "due", shared by the digest, the review issue and the docs queue chart. `spaced-repetition.yml` renders that answer into today's `Spaced Repetition — <Month D, YYYY>` issue — grouped by rung, `WINDOW_CAP` (3) required per rung with the overflow listed as optional, keyed by the ET date in the title so a re-run rewrites one body. It schedules nothing, stores nothing, and writes nothing on Sunday. Without `SRS_ADMIN_KEY` it FAILS rather than inventing a second schedule from issue close dates — that reconstruction was the drift, and it is gone. NEVER add a JSON state file (`.github/problems.json` and friends) or a second "when is this due" rule: D1 answers it, with one writer (`logAttempt`) and git-free history in `attempts`.
+- **The picker owns the local review loop; the buckets are its state.** `work/` holds three buckets (`1` = first solve, `3`/`7` = blind re-solves, named after the ladder rungs) and `bun run pick` has three sections — Tab cycles them. Pools are reconciled from the filesystem plus one `git log` (`apps/cli/work.ts`): a first solve dates from the ET date of the commit that added its `work/1` file — the same push that closes the problem issue and starts the +3/+7 windows — and a window stops being owed the moment its bucket holds a file. Uncapped, unlike the review issue's `WINDOW_CAP`: that issue is one day's assignment, the picker is everything still owed. An empty scaffold under `work/1` owes nothing (it is unsolved), which is why `isImplemented()` lives in `apps/cli/source.ts` and is shared with `close-solved.ts`. leetcode-cli reads its output directory from `~/.leetcode/workspaces/<active>/config.json` and nowhere else — no flag, no env var — so `pick.ts` MERGES `workDir` into that file around the child (never replays saved bytes: the CLI owns the file too) and leaves it pointed at `work/1`, then treats the scaffold's existence as the acceptance test, since leetcode-cli exits 0 even when an expired session made it write nothing. The picker itself writes no state: pushing the re-solve is what moves the ladder, via `close-solved` → `/admin/solved` with that file's bucket.
+- **The README charts are live, never committed.** They are Worker endpoints reading D1 per request (`Cache-Control: max-age=300`, honored by GitHub Camo), so a solved push moves them within ~5 minutes with no commit and no workflow of their own.
+- **D1 owns SRS state; GitHub issues own the catalog; `apps/api/data/schedule.json` owns the calendar.** The catalog reconcile never invents rows and never overwrites SRS columns (`stage`, `next_review`). `Target Date` is the *only* Project field the Worker mirrors (`apps/api/src/mirror.ts`): failures are warnings, never lost D1 writes; topic rows' `Target Date` is never written. SRS stage and first-attempt result live **only** in D1 — do not re-add them as Project fields. The Project's `Set`/`Difficulty` single-selects are a projection of the issue labels, reconciled by `bun run sync-project-fields` — a best-effort final step of `close-solved.yml`, never the Worker.
+- **Determinism = idempotency.** Drill/gate sampling uses a seeded PRNG (`rng()` in `apps/api/src/srs.ts`, FNV-1a → mulberry32, seed = date / ISO week); the digest is keyed by ET date in `email_log`; one-tap links are unique on (problem, date, kind).
+- **DST-proof crons:** each event has two UTC crons; code fires only on the computed ET hour (`etHour`), unit-checked in `apps/api/src/srs.test.ts`.
+- **Chained workflows:** `GITHUB_TOKEN`-driven issue closes fire no `issues` events, so `close-topics.yml` chains off `workflow_run` of Close Solved instead. The Worker's webhook uses its own `GH_PAT`, so its events flow normally.
+
+## Key Directories
+
+| Path | Purpose |
+|---|---|
+| `work/<1\|3\|7>/<Difficulty>/<Category>/<num>.<slug>.{js,py}` | Solutions, e.g. `work/1/Easy/Array/1.two-sum.py`. The top directory is the spaced-repetition bucket: `1` = first solve, `3`/`7` = the blind re-solves of that problem 3 and 7 days later. Machine-parsed header comment (title / `Difficulty:` / URL / `─` rule / statement) — preserve its exact shape |
+| `apps/cli/` | Flat Bun TS workspace: automation entries (shebang + top-level await) and libraries (no shebang, side-effect-free on import). Root `bun run` scripts delegate here |
+| `apps/docs/` | Blume site (`blume.config.ts`, `content/`, `islands/`, `public/`). `content/(<category>)/<num>-<slug>.mdx` generated by sync |
+| `apps/api/` | Cloudflare Worker workspace: `src/` modules, `data/schedule.json` (day → topic issue, human-edited, bundled at deploy), `migrations/` (append-only; `0002` is the +3/+7 ladder rebuild) |
+| `.github/workflows/` | deploy (docs → Pages, Worker → Cloudflare, on every main push), close-solved (issues + `/admin/solved` + Project `Set`/`Difficulty`, needs `SRS_ADMIN_KEY` and `PROJECT_PAT`), close-topics, sync-d1 (issue edits → `/admin/reconcile`), spaced-repetition (daily 06:00 ET cron → today's review issue; `GITHUB_TOKEN` only, no secrets) |
+
+## Development Commands
+
+All from the repo root (a Bun workspace over `apps/*`):
+
+```sh
+bun run pick            # scaffold a solution via leetcode-cli (fuzzy picker; Tab cycles the
+                        # sections: new problems → work/1, then the owed 3-day and 7-day
+                        # re-solves → work/3, work/7. Each section hides what its own bucket
+                        # already holds; index+descriptions cached in
+                        # ~/.local/share/leetcode/data.db — see apps/cli/db.ts)
+bun run stats           # terminal dashboard over GET /api/stats — four panes (Overview,
+                        # Ladder, Concepts, Activity); Tab/1-4 switches, ↑↓ scrolls,
+                        # r refetches, q quits. Outside a TTY it prints every pane once.
+                        # SRS_API points it at a local Worker like close-solved does.
+bun run test            # run ONE solution against LeetCode's judge (not a test suite;
+                        # most recently tested first, stamped in the db on exit 0)
+bun run submit          # submit ONE solution to LeetCode (newest scaffold first, nothing
+                        # hidden; last_submitted stamped in the db on exit 0)
+bun run sync            # work/1 → apps/docs/content/ pages (re-solves are never published)
+bun run dev|build       # Blume docs site (runs in apps/docs/)
+bun run close-solved -- --dry-run    # issue reconcilers (also DRY_RUN=1)
+bun run close-topics -- --dry-run
+# close-solved also pushes the solved set to the Worker; point it at a local
+# one and nothing production is touched:
+SRS_API=http://localhost:8787 SRS_ADMIN_KEY=$LINK_KEY bun run close-solved -- --dry-run
+# Project Set/Difficulty from the issue labels (needs a `project`-scoped token;
+# runs as the last step of close-solved.yml in CI):
+GH_TOKEN=$PROJECT_PAT bun run sync-project-fields -- --dry-run
+# today's review issue from the +3d/+7d closes (--date backfills a missed day;
+# a Sunday date writes nothing — the rest day has no assignment):
+bun run spaced-repetition -- --dry-run --date=2026-08-31
+bun run api:dev         # wrangler dev on :8787 (local D1)
+bun run api:test        # DST guards + the rest-day calendar sweep
+bun run api:deploy      # deploy the Worker by hand (CI also deploys on main pushes)
+curl -X POST -H "Authorization: Bearer $LINK_KEY" \
+  "localhost:8787/admin/digest?dry=1&date=2026-08-30"   # preview a digest, send nothing
+```
+
+## Code Conventions & Common Patterns
+
+- **Scripts are either entries or libraries.** Entries (`apps/cli/close-*.ts`) use shebang + top-level `await`. Libraries (`apps/cli/github.ts`, `apps/cli/report.ts`, everything in `apps/api/src/`) are side-effect-free on import — nothing touches network/credentials until a factory (`github()`, `projectMirror()`) is called.
+- **Shared plumbing:** `github()` gives `repo`, `api()` (throws `METHOD path -> status body`), `list()` (paginating async generator), `closeIssue()` (comment **first**, then close — a failed PATCH still leaves a trace). `report.ts` gives `printTable`/`markdownTable`/`writeStepSummary`.
+- **Narrow API payloads with guards**, not inline casts: `if (!("number" in value) || typeof value.number !== "number") return;` (see `readProblemIssue` in `close-solved.ts`). Named-const casts only with a reason.
+- **Dry-run everything:** reconcilers take `--dry-run`/`DRY_RUN=1`; Worker admin routes take `?dry=1&force=1&date=` (the `SRS_TODAY` equivalent) and `wrangler dev` runs against local D1 — exercise logic there before touching production state.
+- **Dates:** always ET calendar strings (`YYYY-MM-DD`), arithmetic anchored at noon UTC (`atNoon` in `apps/api/src/srs.ts`) to dodge DST. Never `new Date()` math directly.
+- **Style:** section-divider comments (`// ── name ───`), file-top doc comments explaining the *why* and invariants, 2-space JSON with trailing newline, `Map` for dynamic keys / `Record` for static tables, no tiny one-expression wrapper functions.
+- **Known intentional duplication:** `sync.ts` keeps its own header split instead of sharing `apps/cli/source.ts`, because `sync.ts` runs its whole pipeline on import and so cannot be imported from. Don't "deduplicate" it. `close-solved.ts` and `work.ts` do share it — the stub test decides both "close the issue" and "owes a review", and the two must never disagree.
+
+## Important Files
+
+- `apps/api/src/srs.ts` — SRS domain: the ladder (`new → +3 → +7 → retired`; fail resets to `+3`; first-ever log enters at `+3`), `WINDOWS`/`STAGES`/`TEMPERATURES`/`DIFFICULTIES`/`SETS` (the one vocabulary shared by the charts, the `work/<n>` buckets, the review issues and the stats dashboard), ET date math, `workingDay()` (no scheduled date is ever a Sunday), `temperatureOf()` (days since a concept was last exercised → `hot`/`fresh`/`fading`/`cold`, bands derived from `WINDOWS` so they cannot drift; read-side only, nothing is scheduled off it), seeded sampling, `logAttempt()` (the ONE write path — email taps, webhook, gate scoring all converge here), and overload leveling: at most `REVIEW_CAP` (3) reviews surface per day — `levelReviews()` (run by `sendDigest` on real sends only, never dry) gives everything past the cap a concrete future WORKING day, ≤ 3 per day, oldest first, instead of letting the due pile grow.
+- `apps/api/src/index.ts` — router + cron dispatch; `apps/api/wrangler.jsonc` — bindings (`DB`, `EMAIL`), crons, vars; secrets `GH_PAT`/`WEBHOOK_SECRET`/`LINK_KEY` via `wrangler secret put`.
+- `apps/api/src/digest.ts` / `apps/api/src/email.tsx` — the daily digest, split data/presentation. `digest.ts` reads D1 into a `DigestData`; `email.tsx` owns every colour and every sentence, and renders both the HTML and (via its own `plainDigest`, not React Email's `plainText` mode, which flattens the tables) the text alternative. The retrieval rules hold by construction: `DigestRow` has no title and no issue field, so a review or drill line *cannot* leak the topic or a solution link. Subject is `(Day N/56) LeetCode Daily Digest`.
+- `apps/api/src/png.ts` — hand-rolled PNG encoder (RGB8, one IDAT, zlib via `CompressionStream("deflate")`, CRC32, 5×7 bitmap font). It exists because every major email client refuses remote SVG, so `/chart/heatmap.png` rasters the heatmap for the digest while `/chart/heatmap.svg` keeps serving the README byte-for-byte. Both come from one `heatmapCells()` so the two pictures cannot drift.
+- `apps/cli/sync.ts` — work→docs contract: on existing pages only the `## Solution` section itself is script-owned, bounded by `solutionEnd()` at the next unfenced top-level heading; human prose before it and any sections after it (`## Approach`, `## Explanation`) are never touched. Never hand-write solution pages or edit inside `## Solution`.
+- `apps/cli/work.ts` — the `work/` layout: bucket paths, `isSolution()`/`solvedInBucket()` (the stub test, shared with `close-solved.ts`), `firstSolved()` (one `git log`, ET dates, keyed by LC number so past layout moves don't matter), and `reviewQueues()` (what each window still owes). Everything that needs to know where a solution file lives, or when it was solved, goes through here.
+- `apps/api/src/stats.ts` / `apps/cli/stats.ts` / `apps/cli/tui.ts` / `apps/cli/api.ts` — the read side. `buildStats()` is the ONE aggregation: phases, ladder, difficulty, per-topic temperature, gates, the 14-day due queue, and `heat` (every campaign day, zeros and future days included — the docs page windows the tail off it rather than asking a second question). Two readers consume it, the docs island and `bun run stats`, so its shape is both their contract: change all three together. `tui.ts` is hand-rolled ANSI (bars, sparklines, heat cells, alternate screen) on charts.ts's palette, and every chart differs by GLYPH as well as by colour so a piped or `NO_COLOR` dashboard still reads. The TUI recomputes nothing: the only local fact it shows is the `work/` solution count, which is inventory, not schedule.
+- `README.md` live charts are Worker endpoints (`/chart/*.svg`, `/badge/gate.svg`, 5-min Camo cache); the docs `/progress` page fetches `/api/stats` client-side (`apps/docs/islands/ProgressDashboard.tsx`).
+- `apps/docs/blume.config.ts` — site base `/leetcode`; `README.md` campaign table doubles as topic-map input to `sync.ts` (`readmeTopics()`, `TOPIC_ALIASES`).
+- Old `.github/srs/` JSON state, its `srs-*` Actions, and the one-shot `import-srs.ts` importer are RETIRED and deleted — do not resurrect; git history is the record.
+
+## Runtime/Tooling Preferences
+
+- **Bun only**: run scripts with `bun apps/cli/<name>.ts` (or the root `bun run` aliases), install with `bun install --frozen-lockfile`. One workspace (`apps/*`), one root `bun.lock`. Node 22 appears only in `deploy.yml` because Blume requires it.
+- **Near-zero runtime npm dependencies** — SVG, PNG, terminal ANSI, HMAC (WebCrypto), GraphQL are hand-rolled; Actions scripts use Bun builtins + `fetch` only. The **one** exception is the digest email: `apps/api/src/email.tsx` renders React Email (`react`, `react-dom`, `@react-email/components`, `@react-email/render`) inside the Worker, because hand-rolling table-based email HTML that survives Gmail *and* Outlook is not worth owning. It costs ~235 KB gzipped of a 3 MB budget, needs no `nodejs_compat` (it resolves to `renderToReadableStream`), and requires `"jsx": "react-jsx"` in `apps/api/tsconfig.json`. Do not extend this exception to any other module.
+- Local GitHub auth falls back to `gh auth token`; scripts run fine outside Actions.
+
+## Testing & QA
+
+- Repo-wide: no test framework — `bun run test` submits one solution to LeetCode's judge. Exception: `apps/api/src/srs.test.ts` (`bun:test`) proves the DST cron guards with fixed dates; run via `bun run api:test`.
+- QA is dry-runs + local state: Worker logic against `wrangler dev` + local D1 with `?dry=1&date=`; reconcilers with `--dry-run`; idempotency check = run twice, `cmp` output.
+- Docs changes: `bunx blume build --isolated` (from `apps/docs/`) must pass with no new warnings (then `rm -rf apps/docs/.blume-verify`); after sync, review only *created* pages (prose transforms are heuristic).

README.md

@@ -1,111 +1,32 @@
 <p align="center">
-  <img alt="header" src="public/dependency-spine.svg" />
+  <img alt="header" src="https://shieldcn.dev/header/gradient.svg?title=prdlk%2Fleetcode&amp;subtitle=My+Personal+LeetCode+Solutions&amp;logo=leetcode&amp;logoColor=eab308&amp;mode=dark" />
 </p>
 
 <p align="center">
   <a href="https://x.com/basedprad"><img alt="follow" src="https://shieldcn.dev/x/follow/basedprad.svg" /></a>
   <a href="https://lc.prad.nu"><img alt="badge" src="https://shieldcn.dev/badge/Read-Documentation.svg?logo=lu%3ABook" /></a>
+  <a href="https://lc.prad.nu/progress"><img alt="gate pass rate" src="https://srs-api.prdlk.workers.dev/badge/gate.svg" /></a>
 </p>
 
+<p align="center">
+  <sub>8-week interview-prep fight camp. Each phase is a milestone; each topic an issue; each problem a sub-issue.<br>
+  New topic Mon–Fri, gate on Saturday, Sunday off — every solve comes back at <strong>+3</strong> and <strong>+7</strong> days,
+  the two windows that keep the rest day clear (a Thursday solve's +3 slides to Monday).<br>
+  Charts are live from the SRS Worker, never committed: pushing a solution or tapping the daily digest moves them within ~5 min.</sub>
+</p>
 
-## Campaign
-
-<sub>8-week interview-prep fight camp. Each phase is a milestone; each topic an issue; each problem a sub-issue. Core count reflects closed sub-issues.</sub>
-
-<table>
-  <thead>
-    <tr>
-      <th align="center">Phase</th>
-      <th align="center">Wk</th>
-      <th align="left">Date</th>
-      <th align="left">Topic</th>
-      <th align="center">Core</th>
-      <th align="left">Core&nbsp;problems</th>
-    </tr>
-  </thead>
-  <tbody>
-    <tr>
-      <td rowspan="7" align="center"><strong>I</strong><br><sub>Linear Structures</sub></td>
-      <td align="center">1</td><td>Mon Aug 17</td><td><a href="https://github.com/prdlk/leetcode/issues/3"><strong>01</strong> · Hash-Based Lookup</a></td><td align="center">5/5</td><td><a href="https://leetcode.com/problems/contains-duplicate/">217</a> · <a href="https://leetcode.com/problems/valid-anagram/">242</a> · <a href="https://leetcode.com/problems/two-sum/">1</a> · <a href="https://leetcode.com/problems/group-anagrams/">49</a> · <a href="https://leetcode.com/problems/top-k-frequent-elements/">347</a></td>
-    </tr>
-    <tr>
-      <td align="center">1</td><td>Tue Aug 18</td><td><a href="https://github.com/prdlk/leetcode/issues/4"><strong>02</strong> · Prefix Sum</a></td><td align="center">3/3</td><td><a href="https://leetcode.com/problems/range-sum-query-immutable/">303</a> · <a href="https://leetcode.com/problems/product-of-array-except-self/">238</a> · <a href="https://leetcode.com/problems/subarray-sum-equals-k/">560</a></td>
-    </tr>
-    <tr>
-      <td align="center">2</td><td>Mon Aug 24</td><td><a href="https://github.com/prdlk/leetcode/issues/5"><strong>03</strong> · Two Pointers</a></td><td align="center">2/4</td><td><a href="https://leetcode.com/problems/valid-palindrome/">125</a> · <a href="https://leetcode.com/problems/two-sum-ii-input-array-is-sorted/">167</a> · <a href="https://leetcode.com/problems/3sum/">15</a> · <a href="https://leetcode.com/problems/container-with-most-water/">11</a></td>
-    </tr>
-    <tr>
-      <td align="center">2</td><td>Tue Aug 25</td><td><a href="https://github.com/prdlk/leetcode/issues/6"><strong>04</strong> · Sliding Window</a></td><td align="center">1/4</td><td><a href="https://leetcode.com/problems/best-time-to-buy-and-sell-stock/">121</a> · <a href="https://leetcode.com/problems/longest-substring-without-repeating-characters/">3</a> · <a href="https://leetcode.com/problems/longest-repeating-character-replacement/">424</a> · <a href="https://leetcode.com/problems/permutation-in-string/">567</a></td>
-    </tr>
-    <tr>
-      <td align="center">2</td><td>Wed Aug 26</td><td><a href="https://github.com/prdlk/leetcode/issues/7"><strong>05</strong> · Binary Search</a></td><td align="center">3/5</td><td><a href="https://leetcode.com/problems/binary-search/">704</a> · <a href="https://leetcode.com/problems/koko-eating-bananas/">875</a> · <a href="https://leetcode.com/problems/search-in-rotated-sorted-array/">33</a> · <a href="https://leetcode.com/problems/search-a-2d-matrix/">74</a> · <a href="https://leetcode.com/problems/find-minimum-in-rotated-sorted-array/">153</a></td>
-    </tr>
-    <tr>
-      <td align="center">2</td><td>Thu Aug 27</td><td><a href="https://github.com/prdlk/leetcode/issues/8"><strong>06</strong> · Stack</a></td><td align="center">0/3</td><td><a href="https://leetcode.com/problems/valid-parentheses/">20</a> · <a href="https://leetcode.com/problems/min-stack/">155</a> · <a href="https://leetcode.com/problems/evaluate-reverse-polish-notation/">150</a></td>
-    </tr>
-    <tr>
-      <td align="center">2</td><td>Fri Aug 28</td><td><a href="https://github.com/prdlk/leetcode/issues/9"><strong>07</strong> · Monotonic Stack</a></td><td align="center">0/2</td><td><a href="https://leetcode.com/problems/daily-temperatures/">739</a> · <a href="https://leetcode.com/problems/car-fleet/">853</a></td>
-    </tr>
-    <tr>
-      <td rowspan="3" align="center"><strong>II</strong><br><sub>Nodal & Grid</sub></td>
-      <td align="center">3</td><td>Mon Aug 31</td><td><a href="https://github.com/prdlk/leetcode/issues/10"><strong>08</strong> · Linked List</a></td><td align="center">0/3</td><td><a href="https://leetcode.com/problems/reverse-linked-list/">206</a> · <a href="https://leetcode.com/problems/merge-two-sorted-lists/">21</a> · <a href="https://leetcode.com/problems/linked-list-cycle/">141</a></td>
-    </tr>
-    <tr>
-      <td align="center">3</td><td>Tue Sep 1</td><td><a href="https://github.com/prdlk/leetcode/issues/11"><strong>09</strong> · Hybrid Structures</a></td><td align="center">0/2</td><td><a href="https://leetcode.com/problems/lru-cache/">146</a> · <a href="https://leetcode.com/problems/add-two-numbers/">2</a></td>
-    </tr>
-    <tr>
-      <td align="center">3</td><td>Wed Sep 2</td><td><a href="https://github.com/prdlk/leetcode/issues/12"><strong>10</strong> · Matrix Index Math</a></td><td align="center">0/3</td><td><a href="https://leetcode.com/problems/rotate-image/">48</a> · <a href="https://leetcode.com/problems/spiral-matrix/">54</a> · <a href="https://leetcode.com/problems/set-matrix-zeroes/">73</a></td>
-    </tr>
-    <tr>
-      <td rowspan="5" align="center"><strong>III</strong><br><sub>Hierarchical</sub></td>
-      <td align="center">3</td><td>Thu Sep 3</td><td><a href="https://github.com/prdlk/leetcode/issues/13"><strong>11</strong> · Tree DFS</a></td><td align="center">0/3</td><td><a href="https://leetcode.com/problems/invert-binary-tree/">226</a> · <a href="https://leetcode.com/problems/maximum-depth-of-binary-tree/">104</a> · <a href="https://leetcode.com/problems/lowest-common-ancestor-of-a-binary-tree/">236</a></td>
-    </tr>
-    <tr>
-      <td align="center">3</td><td>Fri Sep 4</td><td><a href="https://github.com/prdlk/leetcode/issues/14"><strong>12</strong> · Tree BFS</a></td><td align="center">0/3</td><td><a href="https://leetcode.com/problems/binary-tree-level-order-traversal/">102</a> · <a href="https://leetcode.com/problems/binary-tree-right-side-view/">199</a> · <a href="https://leetcode.com/problems/count-good-nodes-in-binary-tree/">1448</a></td>
-    </tr>
-    <tr>
-      <td align="center">4</td><td>Mon Sep 7</td><td><a href="https://github.com/prdlk/leetcode/issues/15"><strong>13</strong> · Binary Search Tree</a></td><td align="center">0/3</td><td><a href="https://leetcode.com/problems/validate-binary-search-tree/">98</a> · <a href="https://leetcode.com/problems/kth-smallest-element-in-a-bst/">230</a> · <a href="https://leetcode.com/problems/lowest-common-ancestor-of-a-binary-search-tree/">235</a></td>
-    </tr>
-    <tr>
-      <td align="center">4</td><td>Tue Sep 8</td><td><a href="https://github.com/prdlk/leetcode/issues/16"><strong>14</strong> · Heap / Priority Queue</a></td><td align="center">0/3</td><td><a href="https://leetcode.com/problems/kth-largest-element-in-an-array/">215</a> · <a href="https://leetcode.com/problems/k-closest-points-to-origin/">973</a> · <a href="https://leetcode.com/problems/merge-k-sorted-lists/">23</a></td>
-    </tr>
-    <tr>
-      <td align="center">4</td><td>Wed Sep 9</td><td><a href="https://github.com/prdlk/leetcode/issues/17"><strong>15</strong> · Trie</a></td><td align="center">0/2</td><td><a href="https://leetcode.com/problems/implement-trie-prefix-tree/">208</a> · <a href="https://leetcode.com/problems/design-add-and-search-words-data-structure/">211</a></td>
-    </tr>
-    <tr>
-      <td rowspan="4" align="center"><strong>IV</strong><br><sub>Relational</sub></td>
-      <td align="center">4</td><td>Thu Sep 10</td><td><a href="https://github.com/prdlk/leetcode/issues/18"><strong>16</strong> · Graph DFS</a></td><td align="center">0/3</td><td><a href="https://leetcode.com/problems/number-of-islands/">200</a> · <a href="https://leetcode.com/problems/clone-graph/">133</a> · <a href="https://leetcode.com/problems/pacific-atlantic-water-flow/">417</a></td>
-    </tr>
-    <tr>
-      <td align="center">4</td><td>Fri Sep 11</td><td><a href="https://github.com/prdlk/leetcode/issues/19"><strong>17</strong> · Graph BFS</a></td><td align="center">0/2</td><td><a href="https://leetcode.com/problems/rotting-oranges/">994</a> · <a href="https://leetcode.com/problems/word-ladder/">127</a></td>
-    </tr>
-    <tr>
-      <td align="center">5</td><td>Mon Sep 14</td><td><a href="https://github.com/prdlk/leetcode/issues/20"><strong>18</strong> · Topological Sort</a></td><td align="center">0/2</td><td><a href="https://leetcode.com/problems/course-schedule/">207</a> · <a href="https://leetcode.com/problems/course-schedule-ii/">210</a></td>
-    </tr>
-    <tr>
-      <td align="center">5</td><td>Tue Sep 15</td><td><a href="https://github.com/prdlk/leetcode/issues/21"><strong>19</strong> · Union-Find</a></td><td align="center">0/2</td><td><a href="https://leetcode.com/problems/redundant-connection/">684</a> · <a href="https://leetcode.com/problems/number-of-connected-components-in-an-undirected-graph/">323</a></td>
-    </tr>
-    <tr>
-      <td rowspan="5" align="center"><strong>V</strong><br><sub>Decision Space</sub></td>
-      <td align="center">5</td><td>Wed Sep 16</td><td><a href="https://github.com/prdlk/leetcode/issues/22"><strong>20</strong> · Backtracking</a></td><td align="center">0/3</td><td><a href="https://leetcode.com/problems/subsets/">78</a> · <a href="https://leetcode.com/problems/combination-sum/">39</a> · <a href="https://leetcode.com/problems/permutations/">46</a></td>
-    </tr>
-    <tr>
-      <td align="center">5</td><td>Thu Sep 17</td><td><a href="https://github.com/prdlk/leetcode/issues/23"><strong>21</strong> · 1-D Dynamic Programming</a></td><td align="center">0/3</td><td><a href="https://leetcode.com/problems/climbing-stairs/">70</a> · <a href="https://leetcode.com/problems/house-robber/">198</a> · <a href="https://leetcode.com/problems/coin-change/">322</a></td>
-    </tr>
-    <tr>
-      <td align="center">5</td><td>Fri Sep 18</td><td><a href="https://github.com/prdlk/leetcode/issues/24"><strong>22</strong> · Multi-D / Grid DP</a></td><td align="center">0/3</td><td><a href="https://leetcode.com/problems/unique-paths/">62</a> · <a href="https://leetcode.com/problems/minimum-path-sum/">64</a> · <a href="https://leetcode.com/problems/longest-common-subsequence/">1143</a></td>
-    </tr>
-    <tr>
-      <td align="center">6</td><td>Mon Sep 21</td><td><a href="https://github.com/prdlk/leetcode/issues/25"><strong>23</strong> · Greedy</a></td><td align="center">0/3</td><td><a href="https://leetcode.com/problems/maximum-subarray/">53</a> · <a href="https://leetcode.com/problems/jump-game/">55</a> · <a href="https://leetcode.com/problems/gas-station/">134</a></td>
-    </tr>
-    <tr>
-      <td align="center">6</td><td>Tue Sep 22</td><td><a href="https://github.com/prdlk/leetcode/issues/26"><strong>24</strong> · Intervals</a></td><td align="center">0/3</td><td><a href="https://leetcode.com/problems/insert-interval/">57</a> · <a href="https://leetcode.com/problems/merge-intervals/">56</a> · <a href="https://leetcode.com/problems/non-overlapping-intervals/">435</a></td>
-    </tr>
-  </tbody>
+<table align="center">
+  <tr>
+    <td colspan="2" align="center"><img alt="phase progress" width="100%" src="https://srs-api.prdlk.workers.dev/chart/progress.svg" /></td>
+  </tr>
+  <tr>
+    <td align="center" width="53%"><img alt="SRS ladder — new / +3 / +7 / retired" width="100%" src="https://srs-api.prdlk.workers.dev/chart/ladder.svg" /></td>
+    <td align="center" width="47%"><img alt="attempt heatmap" width="100%" src="https://srs-api.prdlk.workers.dev/chart/heatmap.svg" /></td>
+  </tr>
 </table>
 
-## Phase I — Linear Structures (Week 1)
- 
+## Phase I — Linear Structures · Weeks 1–2
+
 <table>
   <thead>
     <tr>
@@ -117,7 +38,7 @@
   </thead>
   <tbody>
     <tr>
-      <td rowspan="9"><strong>1. Hash-Based Lookup</strong><br><sub>builds on nothing</sub></td>
+      <td rowspan="9"><a href="https://github.com/prdlk/leetcode/issues/3"><strong>1. Hash-Based Lookup</strong></a><br><sub>Wk 1 · Mon Aug 17</sub><br><sub>builds on nothing</sub></td>
       <td align="center"><code>A</code></td><td align="center">217</td><td><a href="https://leetcode.com/problems/contains-duplicate/">Contains Duplicate</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">242</td><td><a href="https://leetcode.com/problems/valid-anagram/">Valid Anagram</a></td></tr>
@@ -129,7 +50,7 @@
     <tr><td align="center"><code>B</code></td><td align="center">128</td><td><a href="https://leetcode.com/problems/longest-consecutive-sequence/">Longest Consecutive Sequence</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">290</td><td><a href="https://leetcode.com/problems/word-pattern/">Word Pattern</a></td></tr>
     <tr>
-      <td rowspan="5"><strong>2. Prefix Sum</strong><br><sub>builds on 1</sub></td>
+      <td rowspan="5"><a href="https://github.com/prdlk/leetcode/issues/4"><strong>2. Prefix Sum</strong></a><br><sub>Wk 1 · Tue Aug 18</sub><br><sub>builds on 1</sub></td>
       <td align="center"><code>A</code></td><td align="center">303</td><td><a href="https://leetcode.com/problems/range-sum-query-immutable/">Range Sum Query</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">238</td><td><a href="https://leetcode.com/problems/product-of-array-except-self/">Product of Array Except Self</a></td></tr>
@@ -137,7 +58,7 @@
     <tr><td align="center"><code>B</code></td><td align="center">525</td><td><a href="https://leetcode.com/problems/contiguous-array/">Contiguous Array</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">974</td><td><a href="https://leetcode.com/problems/subarray-sums-divisible-by-k/">Subarray Sums Divisible by K</a></td></tr>
     <tr>
-      <td rowspan="9"><strong>3. Two Pointers</strong><br><sub>builds on ordered iteration</sub></td>
+      <td rowspan="9"><a href="https://github.com/prdlk/leetcode/issues/5"><strong>3. Two Pointers</strong></a><br><sub>Wk 2 · Mon Aug 24</sub><br><sub>builds on ordered iteration</sub></td>
       <td align="center"><code>A</code></td><td align="center">125</td><td><a href="https://leetcode.com/problems/valid-palindrome/">Valid Palindrome</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">167</td><td><a href="https://leetcode.com/problems/two-sum-ii-input-array-is-sorted/">Two Sum II</a></td></tr>
@@ -149,7 +70,7 @@
     <tr><td align="center"><code>B</code></td><td align="center">189</td><td><a href="https://leetcode.com/problems/rotate-array/">Rotate Array</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">80</td><td><a href="https://leetcode.com/problems/remove-duplicates-from-sorted-array-ii/">Remove Duplicates II</a></td></tr>
     <tr>
-      <td rowspan="8"><strong>4. Sliding Window</strong><br><sub>builds on 3</sub></td>
+      <td rowspan="8"><a href="https://github.com/prdlk/leetcode/issues/6"><strong>4. Sliding Window</strong></a><br><sub>Wk 2 · Tue Aug 25</sub><br><sub>builds on 3</sub></td>
       <td align="center"><code>A</code></td><td align="center">121</td><td><a href="https://leetcode.com/problems/best-time-to-buy-and-sell-stock/">Best Time to Buy and Sell Stock</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">3</td><td><a href="https://leetcode.com/problems/longest-substring-without-repeating-characters/">Longest Substring Without Repeating Characters</a></td></tr>
@@ -160,7 +81,7 @@
     <tr><td align="center"><code>B</code></td><td align="center">1004</td><td><a href="https://leetcode.com/problems/max-consecutive-ones-iii/">Max Consecutive Ones III</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">643</td><td><a href="https://leetcode.com/problems/maximum-average-subarray-i/">Maximum Average Subarray</a></td></tr>
     <tr>
-      <td rowspan="9"><strong>5. Binary Search</strong><br><sub>builds on ordered / monotonic invariants</sub></td>
+      <td rowspan="9"><a href="https://github.com/prdlk/leetcode/issues/7"><strong>5. Binary Search</strong></a><br><sub>Wk 2 · Wed Aug 26</sub><br><sub>builds on ordered / monotonic invariants</sub></td>
       <td align="center"><code>A</code></td><td align="center">704</td><td><a href="https://leetcode.com/problems/binary-search/">Binary Search</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">74</td><td><a href="https://leetcode.com/problems/search-a-2d-matrix/">Search a 2D Matrix</a></td></tr>
@@ -172,7 +93,7 @@
     <tr><td align="center"><code>B</code></td><td align="center">34</td><td><a href="https://leetcode.com/problems/find-first-and-last-position-of-element-in-sorted-array/">First &amp; Last Position</a></td></tr>
     <tr><td align="center"><code>B</code> ⭐</td><td align="center">4</td><td><a href="https://leetcode.com/problems/median-of-two-sorted-arrays/">Median of Two Sorted Arrays</a></td></tr>
     <tr>
-      <td rowspan="5"><strong>6. Stack</strong><br><sub>independent primitive</sub></td>
+      <td rowspan="5"><a href="https://github.com/prdlk/leetcode/issues/8"><strong>6. Stack</strong></a><br><sub>Wk 2 · Thu Aug 27</sub><br><sub>independent primitive</sub></td>
       <td align="center"><code>A</code></td><td align="center">20</td><td><a href="https://leetcode.com/problems/valid-parentheses/">Valid Parentheses</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">155</td><td><a href="https://leetcode.com/problems/min-stack/">Min Stack</a></td></tr>
@@ -180,7 +101,7 @@
     <tr><td align="center"><code>B</code></td><td align="center">71</td><td><a href="https://leetcode.com/problems/simplify-path/">Simplify Path</a></td></tr>
     <tr><td align="center"><code>B</code> ⭐</td><td align="center">224</td><td><a href="https://leetcode.com/problems/basic-calculator/">Basic Calculator</a></td></tr>
     <tr>
-      <td rowspan="5"><strong>7. Monotonic Stack</strong><br><sub>builds on 6</sub></td>
+      <td rowspan="5"><a href="https://github.com/prdlk/leetcode/issues/9"><strong>7. Monotonic Stack</strong></a><br><sub>Wk 2 · Fri Aug 28</sub><br><sub>builds on 6</sub></td>
       <td align="center"><code>A</code></td><td align="center">739</td><td><a href="https://leetcode.com/problems/daily-temperatures/">Daily Temperatures</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">853</td><td><a href="https://leetcode.com/problems/car-fleet/">Car Fleet</a></td></tr>
@@ -189,10 +110,10 @@
     <tr><td align="center"><code>B</code> ⭐</td><td align="center">84</td><td><a href="https://leetcode.com/problems/largest-rectangle-in-histogram/">Largest Rectangle in Histogram</a></td></tr>
   </tbody>
 </table>
----
- 
-## Phase II — Nodal &amp; Grid Structures (Week 2)
- 
+
+
+## Phase II — Nodal &amp; Grid Structures · Week 3
+
 <table>
   <thead>
     <tr>
@@ -204,7 +125,7 @@
   </thead>
   <tbody>
     <tr>
-      <td rowspan="10"><strong>8. Linked List</strong><br><sub>independent primitive</sub></td>
+      <td rowspan="10"><a href="https://github.com/prdlk/leetcode/issues/10"><strong>8. Linked List</strong></a><br><sub>Wk 3 · Mon Aug 31</sub><br><sub>independent primitive</sub></td>
       <td align="center"><code>A</code></td><td align="center">206</td><td><a href="https://leetcode.com/problems/reverse-linked-list/">Reverse Linked List</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">21</td><td><a href="https://leetcode.com/problems/merge-two-sorted-lists/">Merge Two Sorted Lists</a></td></tr>
@@ -217,14 +138,14 @@
     <tr><td align="center"><code>B</code></td><td align="center">61</td><td><a href="https://leetcode.com/problems/rotate-list/">Rotate List</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">86</td><td><a href="https://leetcode.com/problems/partition-list/">Partition List</a></td></tr>
     <tr>
-      <td rowspan="4"><strong>9. Hybrid Structures</strong><br><sub>builds on 1 + 8</sub></td>
+      <td rowspan="4"><a href="https://github.com/prdlk/leetcode/issues/11"><strong>9. Hybrid Structures</strong></a><br><sub>Wk 3 · Tue Sep 1</sub><br><sub>builds on 1 + 8</sub></td>
       <td align="center"><code>A</code></td><td align="center">146</td><td><a href="https://leetcode.com/problems/lru-cache/">LRU Cache</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">2</td><td><a href="https://leetcode.com/problems/add-two-numbers/">Add Two Numbers</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">138</td><td><a href="https://leetcode.com/problems/copy-list-with-random-pointer/">Copy List with Random Pointer</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">380</td><td><a href="https://leetcode.com/problems/insert-delete-getrandom-o1/">Insert Delete GetRandom O(1)</a></td></tr>
     <tr>
-      <td rowspan="5"><strong>10. Matrix Index Math</strong><br><sub>builds on 3 — NOT graph traversal</sub></td>
+      <td rowspan="5"><a href="https://github.com/prdlk/leetcode/issues/12"><strong>10. Matrix Index Math</strong></a><br><sub>Wk 3 · Wed Sep 2</sub><br><sub>builds on 3 — NOT graph traversal</sub></td>
       <td align="center"><code>A</code></td><td align="center">48</td><td><a href="https://leetcode.com/problems/rotate-image/">Rotate Image</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">54</td><td><a href="https://leetcode.com/problems/spiral-matrix/">Spiral Matrix</a></td></tr>
@@ -233,10 +154,10 @@
     <tr><td align="center"><code>B</code></td><td align="center">289</td><td><a href="https://leetcode.com/problems/game-of-life/">Game of Life</a></td></tr>
   </tbody>
 </table>
----
- 
-## Phase III — Hierarchical Structures (Week 3)
- 
+
+
+## Phase III — Hierarchical Structures · Weeks 3–4
+
 <table>
   <thead>
     <tr>
@@ -248,7 +169,7 @@
   </thead>
   <tbody>
     <tr>
-      <td rowspan="10"><strong>11. Tree DFS</strong><br><sub>builds on recursion</sub></td>
+      <td rowspan="10"><a href="https://github.com/prdlk/leetcode/issues/13"><strong>11. Tree DFS</strong></a><br><sub>Wk 3 · Thu Sep 3</sub><br><sub>builds on recursion</sub></td>
       <td align="center"><code>A</code></td><td align="center">226</td><td><a href="https://leetcode.com/problems/invert-binary-tree/">Invert Binary Tree</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">104</td><td><a href="https://leetcode.com/problems/maximum-depth-of-binary-tree/">Maximum Depth of Binary Tree</a></td></tr>
@@ -261,7 +182,7 @@
     <tr><td align="center"><code>B</code></td><td align="center">129</td><td><a href="https://leetcode.com/problems/sum-root-to-leaf-numbers/">Sum Root to Leaf Numbers</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">543</td><td><a href="https://leetcode.com/problems/diameter-of-binary-tree/">Diameter of Binary Tree</a></td></tr>
     <tr>
-      <td rowspan="5"><strong>12. Tree BFS</strong><br><sub>builds on 6's sibling (queue) + 11</sub></td>
+      <td rowspan="5"><a href="https://github.com/prdlk/leetcode/issues/14"><strong>12. Tree BFS</strong></a><br><sub>Wk 3 · Fri Sep 4</sub><br><sub>builds on 6's sibling (queue) + 11</sub></td>
       <td align="center"><code>A</code></td><td align="center">102</td><td><a href="https://leetcode.com/problems/binary-tree-level-order-traversal/">Level Order Traversal</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">199</td><td><a href="https://leetcode.com/problems/binary-tree-right-side-view/">Right Side View</a></td></tr>
@@ -269,7 +190,7 @@
     <tr><td align="center"><code>B</code></td><td align="center">103</td><td><a href="https://leetcode.com/problems/binary-tree-zigzag-level-order-traversal/">Zigzag Level Order</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">637</td><td><a href="https://leetcode.com/problems/average-of-levels-in-binary-tree/">Average of Levels</a></td></tr>
     <tr>
-      <td rowspan="7"><strong>13. Binary Search Tree</strong><br><sub>builds on 5 + 11</sub></td>
+      <td rowspan="7"><a href="https://github.com/prdlk/leetcode/issues/15"><strong>13. Binary Search Tree</strong></a><br><sub>Wk 4 · Mon Sep 7</sub><br><sub>builds on 5 + 11</sub></td>
       <td align="center"><code>A</code></td><td align="center">98</td><td><a href="https://leetcode.com/problems/validate-binary-search-tree/">Validate BST</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">230</td><td><a href="https://leetcode.com/problems/kth-smallest-element-in-a-bst/">Kth Smallest in BST</a></td></tr>
@@ -279,7 +200,7 @@
     <tr><td align="center"><code>B</code></td><td align="center">173</td><td><a href="https://leetcode.com/problems/binary-search-tree-iterator/">BST Iterator</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">108</td><td><a href="https://leetcode.com/problems/convert-sorted-array-to-binary-search-tree/">Sorted Array to BST</a></td></tr>
     <tr>
-      <td rowspan="9"><strong>14. Heap / Priority Queue</strong><br><sub>implicit tree over array</sub></td>
+      <td rowspan="9"><a href="https://github.com/prdlk/leetcode/issues/16"><strong>14. Heap / Priority Queue</strong></a><br><sub>Wk 4 · Tue Sep 8</sub><br><sub>implicit tree over array</sub></td>
       <td align="center"><code>A</code></td><td align="center">215</td><td><a href="https://leetcode.com/problems/kth-largest-element-in-an-array/">Kth Largest in Array</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">703</td><td><a href="https://leetcode.com/problems/kth-largest-element-in-a-stream/">Kth Largest in Stream</a></td></tr>
@@ -291,17 +212,17 @@
     <tr><td align="center"><code>B</code> ⭐</td><td align="center">295</td><td><a href="https://leetcode.com/problems/find-median-from-data-stream/">Find Median from Data Stream</a></td></tr>
     <tr><td align="center"><code>B</code> ⭐</td><td align="center">502</td><td><a href="https://leetcode.com/problems/ipo/">IPO</a></td></tr>
     <tr>
-      <td rowspan="3"><strong>15. Trie</strong><br><sub>builds on 11</sub></td>
+      <td rowspan="3"><a href="https://github.com/prdlk/leetcode/issues/17"><strong>15. Trie</strong></a><br><sub>Wk 4 · Wed Sep 9</sub><br><sub>builds on 11</sub></td>
       <td align="center"><code>A</code></td><td align="center">208</td><td><a href="https://leetcode.com/problems/implement-trie-prefix-tree/">Implement Trie</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">211</td><td><a href="https://leetcode.com/problems/design-add-and-search-words-data-structure/">Add &amp; Search Words</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">212</td><td><a href="https://leetcode.com/problems/word-search-ii/">Word Search II</a> <sub>(needs topic 20 — schedule after)</sub></td></tr>
   </tbody>
 </table>
----
- 
-## Phase IV — Relational Structures (Week 4)
- 
+
+
+## Phase IV — Relational Structures · Weeks 4–5
+
 <table>
   <thead>
     <tr>
@@ -313,7 +234,7 @@
   </thead>
   <tbody>
     <tr>
-      <td rowspan="7"><strong>16. Graph DFS</strong><br><sub>builds on 11 + visited set</sub></td>
+      <td rowspan="7"><a href="https://github.com/prdlk/leetcode/issues/18"><strong>16. Graph DFS</strong></a><br><sub>Wk 4 · Thu Sep 10</sub><br><sub>builds on 11 + visited set</sub></td>
       <td align="center"><code>A</code></td><td align="center">200</td><td><a href="https://leetcode.com/problems/number-of-islands/">Number of Islands</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">695</td><td><a href="https://leetcode.com/problems/max-area-of-island/">Max Area of Island</a></td></tr>
@@ -323,7 +244,7 @@
     <tr><td align="center"><code>B</code></td><td align="center">399</td><td><a href="https://leetcode.com/problems/evaluate-division/">Evaluate Division</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">261</td><td><a href="https://leetcode.com/problems/graph-valid-tree/">Graph Valid Tree</a></td></tr>
     <tr>
-      <td rowspan="5"><strong>17. Graph BFS</strong><br><sub>builds on 12 + visited set — shortest path unweighted</sub></td>
+      <td rowspan="5"><a href="https://github.com/prdlk/leetcode/issues/19"><strong>17. Graph BFS</strong></a><br><sub>Wk 4 · Fri Sep 11</sub><br><sub>builds on 12 + visited set — shortest path unweighted</sub></td>
       <td align="center"><code>A</code></td><td align="center">994</td><td><a href="https://leetcode.com/problems/rotting-oranges/">Rotting Oranges</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">127</td><td><a href="https://leetcode.com/problems/word-ladder/">Word Ladder</a></td></tr>
@@ -331,14 +252,14 @@
     <tr><td align="center"><code>B</code></td><td align="center">433</td><td><a href="https://leetcode.com/problems/minimum-genetic-mutation/">Minimum Genetic Mutation</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">286</td><td><a href="https://leetcode.com/problems/walls-and-gates/">Walls and Gates</a></td></tr>
     <tr>
-      <td rowspan="4"><strong>18. Topological Sort</strong><br><sub>builds on 16</sub></td>
+      <td rowspan="4"><a href="https://github.com/prdlk/leetcode/issues/20"><strong>18. Topological Sort</strong></a><br><sub>Wk 5 · Mon Sep 14</sub><br><sub>builds on 16</sub></td>
       <td align="center"><code>A</code></td><td align="center">207</td><td><a href="https://leetcode.com/problems/course-schedule/">Course Schedule</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">210</td><td><a href="https://leetcode.com/problems/course-schedule-ii/">Course Schedule II</a></td></tr>
     <tr><td align="center"><code>B</code> ⭐</td><td align="center">269</td><td><a href="https://leetcode.com/problems/alien-dictionary/">Alien Dictionary</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">1462</td><td><a href="https://leetcode.com/problems/course-schedule-iv/">Course Schedule IV</a></td></tr>
     <tr>
-      <td rowspan="4"><strong>19. Union-Find</strong><br><sub>builds on 16's problem class</sub></td>
+      <td rowspan="4"><a href="https://github.com/prdlk/leetcode/issues/21"><strong>19. Union-Find</strong></a><br><sub>Wk 5 · Tue Sep 15</sub><br><sub>builds on 16's problem class</sub></td>
       <td align="center"><code>A</code></td><td align="center">684</td><td><a href="https://leetcode.com/problems/redundant-connection/">Redundant Connection</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">323</td><td><a href="https://leetcode.com/problems/number-of-connected-components-in-an-undirected-graph/">Number of Connected Components</a></td></tr>
@@ -346,10 +267,10 @@
     <tr><td align="center"><code>B</code></td><td align="center">721</td><td><a href="https://leetcode.com/problems/accounts-merge/">Accounts Merge</a></td></tr>
   </tbody>
 </table>
----
- 
-## Phase V — Decision-Space Exploration (Weeks 4–5)
- 
+
+
+## Phase V — Decision-Space Exploration · Weeks 5–6
+
 <table>
   <thead>
     <tr>
@@ -361,7 +282,7 @@
   </thead>
   <tbody>
     <tr>
-      <td rowspan="8"><strong>20. Backtracking</strong><br><sub>builds on 16 — then unlock LC 212 from topic 15</sub></td>
+      <td rowspan="8"><a href="https://github.com/prdlk/leetcode/issues/22"><strong>20. Backtracking</strong></a><br><sub>Wk 5 · Wed Sep 16</sub><br><sub>builds on 16 — then unlock LC 212 from topic 15</sub></td>
       <td align="center"><code>A</code></td><td align="center">78</td><td><a href="https://leetcode.com/problems/subsets/">Subsets</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">39</td><td><a href="https://leetcode.com/problems/combination-sum/">Combination Sum</a></td></tr>
@@ -372,7 +293,7 @@
     <tr><td align="center"><code>B</code></td><td align="center">22</td><td><a href="https://leetcode.com/problems/generate-parentheses/">Generate Parentheses</a></td></tr>
     <tr><td align="center"><code>B</code> ⭐</td><td align="center">52</td><td><a href="https://leetcode.com/problems/n-queens-ii/">N-Queens II</a></td></tr>
     <tr>
-      <td rowspan="9"><strong>21. 1-D Dynamic Programming</strong><br><sub>builds on 20</sub></td>
+      <td rowspan="9"><a href="https://github.com/prdlk/leetcode/issues/23"><strong>21. 1-D Dynamic Programming</strong></a><br><sub>Wk 5 · Thu Sep 17</sub><br><sub>builds on 20</sub></td>
       <td align="center"><code>A</code></td><td align="center">70</td><td><a href="https://leetcode.com/problems/climbing-stairs/">Climbing Stairs</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">746</td><td><a href="https://leetcode.com/problems/min-cost-climbing-stairs/">Min Cost Climbing Stairs</a></td></tr>
@@ -384,7 +305,7 @@
     <tr><td align="center"><code>B</code></td><td align="center">91</td><td><a href="https://leetcode.com/problems/decode-ways/">Decode Ways</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">647</td><td><a href="https://leetcode.com/problems/palindromic-substrings/">Palindromic Substrings</a></td></tr>
     <tr>
-      <td rowspan="8"><strong>22. Multi-D / Grid DP</strong><br><sub>builds on 21</sub></td>
+      <td rowspan="8"><a href="https://github.com/prdlk/leetcode/issues/24"><strong>22. Multi-D / Grid DP</strong></a><br><sub>Wk 5 · Fri Sep 18</sub><br><sub>builds on 21</sub></td>
       <td align="center"><code>A</code></td><td align="center">62</td><td><a href="https://leetcode.com/problems/unique-paths/">Unique Paths</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">64</td><td><a href="https://leetcode.com/problems/minimum-path-sum/">Minimum Path Sum</a></td></tr>
@@ -395,7 +316,7 @@
     <tr><td align="center"><code>B</code></td><td align="center">221</td><td><a href="https://leetcode.com/problems/maximal-square/">Maximal Square</a></td></tr>
     <tr><td align="center"><code>B</code> ⭐</td><td align="center">97</td><td><a href="https://leetcode.com/problems/interleaving-string/">Interleaving String</a></td></tr>
     <tr>
-      <td rowspan="7"><strong>23. Greedy</strong><br><sub>builds on 21 — includes Kadane's</sub></td>
+      <td rowspan="7"><a href="https://github.com/prdlk/leetcode/issues/25"><strong>23. Greedy</strong></a><br><sub>Wk 6 · Mon Sep 21</sub><br><sub>builds on 21 — includes Kadane's</sub></td>
       <td align="center"><code>A</code></td><td align="center">53</td><td><a href="https://leetcode.com/problems/maximum-subarray/">Maximum Subarray</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">55</td><td><a href="https://leetcode.com/problems/jump-game/">Jump Game</a></td></tr>
@@ -405,7 +326,7 @@
     <tr><td align="center"><code>B</code></td><td align="center">918</td><td><a href="https://leetcode.com/problems/maximum-sum-circular-subarray/">Maximum Sum Circular Subarray</a></td></tr>
     <tr><td align="center"><code>B</code></td><td align="center">763</td><td><a href="https://leetcode.com/problems/partition-labels/">Partition Labels</a></td></tr>
     <tr>
-      <td rowspan="7"><strong>24. Intervals</strong><br><sub>builds on 23</sub></td>
+      <td rowspan="7"><a href="https://github.com/prdlk/leetcode/issues/26"><strong>24. Intervals</strong></a><br><sub>Wk 6 · Tue Sep 22</sub><br><sub>builds on 23</sub></td>
       <td align="center"><code>A</code></td><td align="center">57</td><td><a href="https://leetcode.com/problems/insert-interval/">Insert Interval</a></td>
     </tr>
     <tr><td align="center"><code>A</code></td><td align="center">56</td><td><a href="https://leetcode.com/problems/merge-intervals/">Merge Intervals</a></td></tr>

apps/api/.dev.vars.example

@@ -0,0 +1,3 @@
+GH_PAT=github_pat_or_gho_token_with_issues_and_projects_rw
+WEBHOOK_SECRET=random_hex_matching_the_repo_webhook
+LINK_KEY=random_hex_signing_one_tap_links

apps/api/README.md

@@ -0,0 +1,126 @@
+# srs-api
+
+Cloudflare Worker running the spaced-repetition system: daily digest email,
+one-tap logging, `/done` webhook, Saturday review issues, live SVG charts,
+and the `/api/stats` feed for the docs progress page.
+
+Live at `https://srs-api.prdlk.workers.dev`.
+
+## Direction of truth
+
+- **D1 owns SRS state** — stages, review dates, attempt history, gate
+  scores, boost flags. Nothing else is authoritative.
+- **GitHub issues own the catalog** — topics, problems, `set:*`/`diff:*`
+  labels, milestones. `catalog.ts` reconciles them into D1 nightly (and via
+  `/admin/reconcile`); it recomputes from scratch, never invents rows, and
+  never overwrites SRS-owned columns (`stage`, `next_review`).
+- **The repo owns the schedule** — `data/schedule.json`, human-edited,
+  bundled at deploy. Git history is its audit log.
+- The GitHub Project mirror writes `Target Date` and nothing else: it is
+  best-effort, so failures log warnings and never block a D1 write, and topic
+  rows are never written. SRS stage and first-attempt result stay in D1 only —
+  the charts, digest and `/api/stats` read them from there, so mirroring them
+  into Project single-selects bought only drift. Do not re-add them.
+  `Set`/`Difficulty` are a projection of the issue labels, reconciled by
+  `bun run sync-project-fields` at the repo root, not by this Worker.
+
+## The ladder
+
+`new → +3 → +7 → retired`; pass advances, fail resets to `+3`. A problem's
+first-ever log enters at `+3` regardless of result. Stage names the NEXT
+review's interval, and the two rungs are the same numbers as the `work/3` and
+`work/7` buckets `bun run pick` scaffolds re-solves into.
+
+**Sunday is never booked.** Topics run Mon–Fri, the gate is Saturday, and 3/7
+are chosen so a solve returns on a working day: only a Thursday solve's `+3`
+would land on the rest day, and `workingDay()` in `src/srs.ts` slides it to
+Monday. Every scheduled date in the Worker — ladder reviews, levelled
+overflow, deferred-Hard release dates — is minted by that one function, so the
+rest day cannot be booked and then swallowed by the digest's Sunday branch.
+Reviews may slip later, never earlier: pulling one back to Saturday would
+shorten the interval it exists to test. `src/srs.test.ts` sweeps the whole
+campaign calendar to prove it.
+
+All dates are ET calendar strings anchored at noon UTC (`src/srs.ts`) — never
+raw `Date` math.
+
+Blind drills draw only from topics **already learned**: scheduled in an
+earlier week (the current week's optional pool is reserved for Saturday's
+gate) and with at least one core problem on the ladder — a skipped learning
+day never feeds drills just because its calendar week lapsed.
+
+## Routes
+
+| Route | Auth | Purpose |
+|---|---|---|
+| `GET /log?p&r&d&sig` | HMAC (`LINK_KEY`) | one-tap pass/fail; idempotent per (problem, date, kind); links expire after 3 days |
+| `POST /webhook/github` | HMAC (`WEBHOOK_SECRET`) | `/done <n> pass\|fail` comments (owner only, any issue); `review`-issue close → gate scoring |
+| `GET /chart/{progress,ladder,heatmap}.svg`, `GET /badge/gate.svg` | public | hand-rolled SVGs, `max-age=300` (GitHub Camo's freshness floor) |
+| `GET /api/stats` | public, CORS-pinned to the docs origin | one JSON document for `/progress` |
+| `POST /admin/solved` | `Authorization: Bearer <LINK_KEY>` | `{"solved":[{"lc":1,"bucket":1},…]}` — solutions committed under `work/`; the bucket names the rung it settles (1 → `new`, 3 → `+3`, 7 → `+7`), so a first solve enters the ladder and a pushed re-solve advances it. Writes only when the problem stands on that rung, which is what makes re-posting the whole set a no-op. Takes `?dry=1&date=` |
+| `POST /admin/due` | `Authorization: Bearer <LINK_KEY>` | read-only: what the ladder has due on `?date=` (default today), oldest first, with the rung/bucket, how late it is, and when it was last seen. The one answer behind the daily `Spaced Repetition` issue |
+| `POST /admin/{digest,review,reconcile}` | `Authorization: Bearer <LINK_KEY>` | manual triggers; `digest` takes `?dry=1&force=1&date=` |
+
+## Crons (DST-proof)
+
+Each event has two UTC crons; code fires only when the computed ET hour
+matches (`etHour` in `src/srs.ts`, unit-checked in `src/srs.test.ts`):
+
+- `0 12,13 * * *` → 8 AM ET: catalog reconcile, then the digest.
+- `0 4,5 * * 6` → midnight ET Saturday: create `Review — Week N` (so the
+  8 AM digest can link to it).
+
+## Secrets & bindings
+
+`wrangler secret put` — `GH_PAT` (Issues + Projects RW), `WEBHOOK_SECRET`
+(matches the repo webhook), `LINK_KEY` (signs one-tap links, gates admin
+routes). Bindings in `wrangler.jsonc`: `DB` (D1 `srs`), `EMAIL`
+(`send_email`, restricted to the verified destination). Sender domain
+`prdlk.com` is onboarded to Email Sending.
+
+Three Actions workflows talk to this Worker with the `SRS_ADMIN_KEY` repo
+secret:
+
+- `sync-d1.yml` sends curriculum issue edits to `POST /admin/reconcile` (the
+  morning cron is the backstop).
+- `close-solved.yml` sends the whole implemented `work/` set — each entry
+  tagged with its bucket — to `POST /admin/solved` on every push to `main`.
+  That call is what keeps a committed solution from being invisible here:
+  closing an issue is not a state change the Worker can see (the catalog
+  reconcile ignores issue state, and only `logAttempt()` moves the ladder), and
+  the bucket is what lets a pushed re-solve advance `+3 → +7` instead of
+  being read as a stale first solve. `close-solved` still owns the issue close
+  itself (it knows the files and the commit), so this path only writes D1 and
+  mirrors Project fields.
+- `spaced-repetition.yml` reads `POST /admin/due` at 06:00 ET and renders
+  today's `Spaced Repetition` issue. Read-only, and the only schedule involved
+  is this Worker's ladder.
+
+## Local dev
+
+```sh
+cp .dev.vars.example .dev.vars       # or fill GH_PAT/WEBHOOK_SECRET/LINK_KEY
+bun install
+bunx wrangler d1 migrations apply srs --local
+bun run dev                          # wrangler dev on :8787, local D1
+curl -X POST -H "Authorization: Bearer $LINK_KEY" \
+  "localhost:8787/admin/digest?dry=1&date=2026-08-31"   # prints HTML, sends nothing
+bun test src                         # DST guards, date math, the rest-day sweep
+```
+
+`?date=` on admin routes is the `SRS_TODAY` equivalent. Migrations are
+append-only and applied in order; `0002_ladder_3_7.sql` is the +2/+5/+10 →
+`+3`/`+7` rebuild, which also lifts every date that was sitting on a Sunday.
+The one-shot importer for the retired `.github/srs/srs.json` is gone — that
+state no longer exists, and git history is its record.
+
+## Deploy
+
+```sh
+bun run deploy        # from apps/api/, or `bun run api:deploy` at the repo root
+```
+
+CI deploys on every push to `main` (`deploy.yml`, gated on `bun run api:test`,
+using the `CLOUDFLARE_API_TOKEN`/`CLOUDFLARE_ACCOUNT_ID` repo secrets); the
+commands above are for out-of-band deploys. After changing bindings, rerun
+`bunx wrangler types`.

apps/api/data/schedule.json

@@ -0,0 +1,26 @@
+{
+  "2026-08-17": 3,
+  "2026-08-18": 4,
+  "2026-08-24": 5,
+  "2026-08-25": 6,
+  "2026-08-26": 7,
+  "2026-08-27": 8,
+  "2026-08-28": 9,
+  "2026-08-31": 10,
+  "2026-09-01": 11,
+  "2026-09-02": 12,
+  "2026-09-03": 13,
+  "2026-09-04": 14,
+  "2026-09-07": 15,
+  "2026-09-08": 16,
+  "2026-09-09": 17,
+  "2026-09-10": 18,
+  "2026-09-11": 19,
+  "2026-09-14": 20,
+  "2026-09-15": 21,
+  "2026-09-16": 22,
+  "2026-09-17": 23,
+  "2026-09-18": 24,
+  "2026-09-21": 25,
+  "2026-09-22": 26
+}

apps/api/migrations/0001_init.sql

@@ -0,0 +1,55 @@
+-- SRS schema. D1 owns SRS state (stage, next_review, attempts, gates,
+-- boosts); GitHub issues own the catalog columns, reconciled nightly.
+CREATE TABLE problems (
+  lc_number   INTEGER PRIMARY KEY,
+  issue       INTEGER NOT NULL,
+  topic_issue INTEGER NOT NULL,
+  title       TEXT NOT NULL,
+  difficulty  TEXT NOT NULL,               -- easy|medium|hard
+  set_label   TEXT NOT NULL,               -- core|optional|deferred
+  stage       TEXT NOT NULL DEFAULT 'new', -- new|+2|+5|+10|retired
+  next_review TEXT,                        -- YYYY-MM-DD ET, NULL when retired/unsolved
+  defer_until TEXT                         -- deferred Hards: 2026-09-28+
+);
+
+CREATE TABLE attempts (
+  id        INTEGER PRIMARY KEY AUTOINCREMENT,
+  lc_number INTEGER NOT NULL REFERENCES problems(lc_number),
+  date      TEXT NOT NULL,
+  kind      TEXT NOT NULL,                 -- first|review|drill|gate
+  result    TEXT NOT NULL,                 -- pass|fail
+  source    TEXT NOT NULL                  -- email|webhook|commit|import
+);
+
+-- One-tap email links must be idempotent (same link twice = no-op), but
+-- /done webhook corrections (pass then fail, same day) must stay legal —
+-- so uniqueness applies to email-sourced attempts only. Commit-sourced
+-- attempts need no index: logAttempt() only accepts them for a stage-'new'
+-- problem, so a re-push cannot insert a second one.
+CREATE UNIQUE INDEX attempts_email_once
+  ON attempts (lc_number, date, kind) WHERE source = 'email';
+CREATE INDEX attempts_by_date ON attempts (date);
+
+CREATE TABLE topics (
+  issue     INTEGER PRIMARY KEY,
+  name      TEXT NOT NULL,
+  misses    INTEGER NOT NULL DEFAULT 0,
+  boost     INTEGER NOT NULL DEFAULT 0,
+  milestone INTEGER                        -- phase; from the topic issue
+);
+
+CREATE TABLE gates (
+  week      INTEGER PRIMARY KEY,           -- ISO week
+  issue     INTEGER,
+  problems  TEXT NOT NULL,                 -- JSON array of lc_numbers
+  pass_rate REAL,
+  closed_on TEXT
+);
+
+CREATE TABLE drill_pool_used (lc_number INTEGER PRIMARY KEY);
+
+-- Digest idempotency: one email per ET date.
+CREATE TABLE email_log (
+  date    TEXT PRIMARY KEY,                -- YYYY-MM-DD ET
+  sent_at TEXT NOT NULL
+);

apps/api/migrations/0002_ladder_3_7.sql

@@ -0,0 +1,31 @@
+-- Ladder rebuild: +2/+5/+10 → +3/+7, and no scheduled date on a Sunday.
+--
+-- The old three rungs scattered reviews across every weekday, so roughly one
+-- review in seven came due on the rest day: the digest hid it (Sunday builds
+-- an empty body) and Monday inherited the pile. The campaign's windows are 3
+-- and 7 days — the same numbers as the work/3 and work/7 buckets and the
+-- Spaced Repetition issue windows — and srs.ts now mints every date through
+-- workingDay(), which never returns a Sunday.
+--
+-- Stage remap keeps earned progress: a problem at +5 has passed one review, so
+-- it owes the 7-day one; a problem at +10 has passed two and is done. Existing
+-- next_review dates keep their day (recomputing them would move work the
+-- reader has already been told about) — only Sunday landings shift, forward to
+-- Monday, because a review may come later than its interval but never sooner.
+--
+-- problems.stage is now: new | +3 | +7 | retired
+
+UPDATE problems SET stage = '+3' WHERE stage = '+2';
+UPDATE problems SET stage = '+7' WHERE stage = '+5';
+UPDATE problems SET stage = 'retired', next_review = NULL WHERE stage = '+10';
+
+UPDATE problems
+   SET next_review = date(next_review, '+1 day')
+ WHERE next_review IS NOT NULL AND strftime('%w', next_review) = '0';
+
+-- defer_until is the deferred Hards' release date; catalog.ts spreads them two
+-- per working day, so the same rule applies. next_review tracks it for a row
+-- that is still `new`, and the shift above kept the two in step.
+UPDATE problems
+   SET defer_until = date(defer_until, '+1 day')
+ WHERE defer_until IS NOT NULL AND strftime('%w', defer_until) = '0';

apps/api/package.json

@@ -0,0 +1,21 @@
+{
+  "name": "srs-api",
+  "private": true,
+  "type": "module",
+  "scripts": {
+    "dev": "wrangler dev --test-scheduled",
+    "deploy": "wrangler deploy",
+    "types": "wrangler types",
+    "test": "bun test src"
+  },
+  "devDependencies": {
+    "@types/react": "^19.2.18",
+    "wrangler": "^4.125.0"
+  },
+  "dependencies": {
+    "@react-email/components": "^1.0.12",
+    "@react-email/render": "^2.1.0",
+    "react": "^19.2.8",
+    "react-dom": "^19.2.8"
+  }
+}

apps/api/src/catalog.ts

@@ -0,0 +1,120 @@
+/**
+ * Nightly GitHub → D1 catalog reconcile. GitHub issues own the catalog
+ * (topics, problems, set/diff labels, milestones); this recomputes desired
+ * rows from scratch on every run — re-runs are no-ops. SRS-owned columns
+ * (stage, next_review) are NEVER overwritten; defer_until is set once, on
+ * first sight of a deferred problem. The Worker never invents catalog rows.
+ *
+ * Difficulty and set come from the `diff:*` / `set:*` LABELS, never from the
+ * issue title — the title carries only the LC number and the problem's name,
+ * so relabelling a problem is enough to move it and titles stay readable.
+ * The `problem` label is what marks a sub-issue as curriculum.
+ */
+import { DIFFICULTIES, SETS, workingDay } from "./srs.ts";
+import type { GitHub } from "./github.ts";
+
+const TITLE_RE = /^LC (\d+) · (.+)$/;
+
+/** Deferred Hards enter the queue from this date, two per working day. */
+const DEFER_FROM = "2026-09-28";
+
+export interface ReconcileReport {
+  topics: number;
+  problems: number;
+}
+
+/** Label names on one issue payload, narrowed with guards rather than cast. */
+function labelNames(value: object): string[] {
+  if (!("labels" in value) || !Array.isArray(value.labels)) return [];
+  const names: string[] = [];
+  for (const label of value.labels) {
+    if (label && typeof label === "object" && "name" in label && typeof label.name === "string") {
+      names.push(label.name);
+    }
+  }
+  return names;
+}
+
+export async function reconcileCatalog(db: D1Database, gh: GitHub): Promise<ReconcileReport> {
+  interface TopicIssue {
+    number: number;
+    title: string;
+    milestone: number | null;
+  }
+  const topics: TopicIssue[] = [];
+  for await (const raw of gh.list(`/repos/${gh.repo}/issues?labels=topic&state=all`)) {
+    if (!raw || typeof raw !== "object") continue;
+    if ("pull_request" in raw) continue;
+    if (!("number" in raw) || typeof raw.number !== "number") continue;
+    if (!("title" in raw) || typeof raw.title !== "string") continue;
+    let milestone: number | null = null;
+    if (
+      "milestone" in raw &&
+      raw.milestone &&
+      typeof raw.milestone === "object" &&
+      "number" in raw.milestone &&
+      typeof raw.milestone.number === "number"
+    ) {
+      milestone = raw.milestone.number;
+    }
+    topics.push({ number: raw.number, title: raw.title, milestone });
+  }
+
+  const statements: D1PreparedStatement[] = [];
+  for (const t of topics) {
+    statements.push(
+      db
+        .prepare(
+          `INSERT INTO topics (issue, name, milestone) VALUES (?1, ?2, ?3)
+           ON CONFLICT(issue) DO UPDATE SET name = ?2, milestone = ?3`,
+        )
+        .bind(t.number, t.title.replace(/^Topic \d+ — /, ""), t.milestone),
+    );
+  }
+
+  let problems = 0;
+  let deferredSeen = 0;
+  for (const t of topics) {
+    for await (const raw of gh.list(`/repos/${gh.repo}/issues/${t.number}/sub_issues`)) {
+      if (!raw || typeof raw !== "object") continue;
+      if (!("number" in raw) || typeof raw.number !== "number") continue;
+      if (!("title" in raw) || typeof raw.title !== "string") continue;
+      const labels = labelNames(raw);
+      if (!labels.includes("problem")) continue; // non-curriculum sub-issue
+      const m = raw.title.match(TITLE_RE);
+      if (!m) {
+        console.warn(`catalog: #${raw.number} is a problem but not titled "LC <n> · <name>"`);
+        continue;
+      }
+      const difficulty = DIFFICULTIES.find((d) => labels.includes(`diff:${d}`));
+      const set = SETS.find((s) => labels.includes(`set:${s}`));
+      if (!difficulty || !set) {
+        console.warn(`catalog: #${raw.number} is missing a diff:/set: label — skipped`);
+        continue;
+      }
+      const lc = Number(m[1]);
+      // Two deferred Hards per WORKING day from DEFER_FROM, in catalog walk
+      // order — applied only when the row is first created (SRS owns it
+      // afterwards). Sunday is skipped like every other scheduled date.
+      const defer = set === "deferred" ? workingDay(DEFER_FROM, Math.floor(deferredSeen / 2)) : null;
+      if (set === "deferred") deferredSeen++;
+      statements.push(
+        db
+          .prepare(
+            `INSERT INTO problems (lc_number, issue, topic_issue, title, difficulty, set_label, defer_until, next_review)
+             VALUES (?1, ?2, ?3, ?4, ?5, ?6, ?7, ?7)
+             ON CONFLICT(lc_number) DO UPDATE SET
+               issue = ?2, topic_issue = ?3, title = ?4, difficulty = ?5, set_label = ?6`,
+          )
+          .bind(lc, raw.number, t.number, m[2]!, difficulty, set, defer),
+      );
+      problems++;
+    }
+  }
+
+  // D1 batches are transactional; chunk to stay under statement limits.
+  for (let i = 0; i < statements.length; i += 50) {
+    await db.batch(statements.slice(i, i + 50));
+  }
+  return { topics: topics.length, problems };
+}

apps/api/src/charts.ts

@@ -0,0 +1,408 @@
+/**
+ * Hand-rolled SVG chart generation for the SRS Worker.
+ *
+ * Feeds five endpoints — /chart/progress.svg, /chart/ladder.svg,
+ * /chart/heatmap.svg, /chart/heatmap.png, /badge/gate.svg — each served with
+ * `Cache-Control: public, max-age=300`. The heatmap has a raster twin because
+ * every major email client blocks SVG, so the digest cannot embed the vector.
+ *
+ * The SVG is string-built: no chart library, no dependencies, and the one
+ * import is the hand-rolled PNG encoder next door. The only dynamic values
+ * entering the markup are numbers, percentages, and dates the Worker itself
+ * computes, plus the fixed phase/stage labels — so nothing here needs XML
+ * escaping. Styling is shadcn dark zinc to match the README's shieldcn
+ * badges: rounded #09090b cards, #fafafa/#a1a1aa text, GitHub dark-mode
+ * green ramp. Every function renders on a zero-row DB.
+ */
+import { STAGES } from "./srs.ts";
+import { Canvas, textWidth } from "./png.ts";
+
+// D1Database comes from the generated worker-configuration.d.ts runtime types.
+
+const FONT = "Verdana,DejaVu Sans,sans-serif";
+
+// shadcn dark-zinc palette, matching the README's shieldcn badges.
+const BG = "#09090b"; // zinc-950 card
+const FG = "#fafafa"; // zinc-50 text
+const MUTED = "#a1a1aa"; // zinc-400 secondary text
+const LINE = "#27272a"; // zinc-800 structure / empty
+const GREEN = "#16a34a"; // green-600 (core / pass)
+const BLUE = "#2563eb"; // blue-600 (optional)
+const RED = "#f87171"; // red-400 (fail, on dark)
+
+// Green ramp shared by the heatmap and the ladder (GitHub dark-mode
+// contribution hues; level 0 is the empty zinc cell).
+const GREENS = ["#27272a", "#0e4429", "#006d32", "#26a641", "#39d353"];
+
+// Labels are stored SVG-ready: phase II's "&" is pre-escaped since these
+// strings go straight into markup and nothing else here needs escaping.
+const PHASES = [
+  { milestone: 1, name: "I — Linear" },
+  { milestone: 2, name: "II — Nodal & Grid" },
+  { milestone: 3, name: "III — Hierarchical" },
+  { milestone: 4, name: "IV — Relational" },
+  { milestone: 5, name: "V — Decision Space" },
+];
+
+// ── svg helpers ──────────────────────────────────────────────────
+
+/** Opening tag plus the rounded dark card every chart starts with. */
+function svgOpen(width: number, height: number): string {
+  return (
+    `<svg xmlns="http://www.w3.org/2000/svg" width="${width}" height="${height}" ` +
+    `viewBox="0 0 ${width} ${height}" font-family="${FONT}">` +
+    `<rect width="${width}" height="${height}" rx="6" fill="${BG}"/>`
+  );
+}
+
+/** A <text> element; charts place dozens of these with the same defaults. */
+function text(
+  x: number,
+  y: number,
+  content: string,
+  attrs: { size?: number; fill?: string; anchor?: string; weight?: string } = {},
+): string {
+  const { size = 12, fill = FG, anchor = "start", weight } = attrs;
+  const bold = weight ? ` font-weight="${weight}"` : "";
+  return `<text x="${x}" y="${y}" font-size="${size}" fill="${fill}" text-anchor="${anchor}"${bold}>${content}</text>`;
+}
+
+// ── progress: stacked bar per phase ──────────────────────────────
+
+interface ProgressRow {
+  milestone: number;
+  set_label: string;
+  done: number;
+  total: number;
+}
+
+export async function progressChart(db: D1Database): Promise<string> {
+  const { results } = await db
+    .prepare(
+      `SELECT t.milestone AS milestone, p.set_label AS set_label,
+              SUM(CASE WHEN p.stage != 'new' THEN 1 ELSE 0 END) AS done,
+              COUNT(*) AS total
+       FROM problems p JOIN topics t ON p.topic_issue = t.issue
+       WHERE t.milestone IS NOT NULL
+       GROUP BY t.milestone, p.set_label`,
+    )
+    .all<ProgressRow>();
+
+  // Per phase: core done / optional done / remaining (all-set total − all done).
+  const phases = PHASES.map((phase) => {
+    const rows = results.filter((r) => r.milestone === phase.milestone);
+    const total = rows.reduce((n, r) => n + r.total, 0);
+    const done = rows.reduce((n, r) => n + r.done, 0);
+    return {
+      name: phase.name,
+      coreDone: rows.find((r) => r.set_label === "core")?.done ?? 0,
+      optionalDone: rows.find((r) => r.set_label === "optional")?.done ?? 0,
+      remaining: total - done,
+      done,
+      total,
+    };
+  });
+
+  const width = 640;
+  const height = 300;
+  const barX = 168;
+  const barMaxW = 400;
+  const barH = 22;
+  const rowStep = 48;
+  const scale = Math.max(1, ...phases.map((p) => p.total));
+
+  const parts = [svgOpen(width, height)];
+
+  // Legend on top.
+  const legend: [string, string][] = [
+    [GREEN, "core done"],
+    [BLUE, "optional done"],
+    [LINE, "remaining"],
+  ];
+  let lx = barX;
+  for (const [color, label] of legend) {
+    parts.push(`<rect x="${lx}" y="12" width="12" height="12" rx="2" fill="${color}"/>`);
+    parts.push(text(lx + 17, 22, label, { size: 11, fill: MUTED }));
+    lx += 17 + label.length * 7 + 24;
+  }
+
+  phases.forEach((p, i) => {
+    const y = 52 + i * rowStep;
+    const midY = y + barH / 2 + 4;
+    parts.push(text(barX - 12, midY, p.name, { anchor: "end" }));
+
+    let x = barX;
+    const segments: [number, string][] = [
+      [p.coreDone, GREEN],
+      [p.optionalDone, BLUE],
+      [p.remaining, LINE],
+    ];
+    for (const [count, color] of segments) {
+      const w = (count / scale) * barMaxW;
+      if (w > 0) {
+        parts.push(`<rect x="${x.toFixed(1)}" y="${y}" width="${w.toFixed(1)}" height="${barH}" fill="${color}"/>`);
+        // Count inside the segment when it fits; tiny slivers stay unlabeled.
+        if (w >= 20) {
+          const labelFill = color === LINE ? MUTED : FG;
+          parts.push(text(x + w / 2, midY, String(count), { size: 11, fill: labelFill, anchor: "middle" }));
+        }
+        x += w;
+      }
+    }
+    parts.push(text(x + 8, midY, `${p.done}/${p.total}`, { size: 11, fill: MUTED }));
+  });
+
+  parts.push("</svg>");
+  return parts.join("");
+}
+
+// ── ladder: bar per stage ────────────────────────────────────────
+
+// One bar per rung, labelled and ordered by srs.ts's STAGES, so changing the
+// ladder can never leave a stale bar here: the slot width is derived from the
+// rung count, only the bar width inside a slot is fixed.
+export async function ladderChart(db: D1Database): Promise<string> {
+  const { results } = await db
+    .prepare(`SELECT stage, COUNT(*) AS count FROM problems GROUP BY stage`)
+    .all<{ stage: string; count: number }>();
+
+  const counts = STAGES.map((s) => results.find((r) => r.stage === s)?.count ?? 0);
+
+  const width = 640;
+  const height = 220;
+  const plotTop = 26;
+  const plotBottom = height - 32;
+  const plotH = plotBottom - plotTop;
+  const slotW = (width - 80) / STAGES.length;
+  const barW = 64;
+  const scale = Math.max(1, ...counts);
+
+  const parts = [svgOpen(width, height)];
+
+  counts.forEach((count, i) => {
+    const cx = 40 + slotW * i + slotW / 2;
+    const barH = (count / scale) * plotH;
+    const y = plotBottom - barH;
+    if (barH > 0) {
+      parts.push(
+        `<rect x="${(cx - barW / 2).toFixed(1)}" y="${y.toFixed(1)}" width="${barW}" height="${barH.toFixed(1)}" rx="3" fill="${GREENS[i]}"/>`,
+      );
+    }
+    parts.push(text(cx, Math.max(y - 6, 16), String(count), { size: 12, anchor: "middle", weight: "bold" }));
+    parts.push(text(cx, plotBottom + 18, STAGES[i]!, { size: 12, fill: MUTED, anchor: "middle" }));
+  });
+
+  parts.push(`<line x1="40" y1="${plotBottom}" x2="${width - 40}" y2="${plotBottom}" stroke="${LINE}"/>`);
+  parts.push("</svg>");
+  return parts.join("");
+}
+
+// ── heatmap: attempts per campaign day ───────────────────────────
+
+const DAY_MS = 86_400_000;
+
+// Layout in CSS pixels. heatmapPng multiplies every one of these by its device
+// scale, so the vector and raster pictures cannot drift apart.
+const HEAT_W = 560;
+const HEAT_H = 160;
+const HEAT_GRID_X = 34;
+const HEAT_GRID_Y = 24;
+const HEAT_CELL = 16;
+const HEAT_STEP = 19; // cell + 3px gap
+
+// Grid row → left-hand label; both renderers print only these three.
+const WEEKDAYS: [number, string][] = [
+  [0, "Mon"],
+  [2, "Wed"],
+  [4, "Fri"],
+];
+
+/** One grid square: column, Mon-based row, and index into GREENS. */
+interface HeatmapCell {
+  col: number;
+  row: number;
+  level: number;
+}
+
+/**
+ * The heatmap's whole data model — one cell per campaign day, plus the week
+ * count that positions the legend — shared by both renderers.
+ *
+ * Every per-day date derives from one UTC-midnight timestamp, so local-timezone
+ * drift never shifts a cell. The campaign starts on a Monday, so day i sits at
+ * column i/7; the row comes from the real weekday, so an off-Monday start still
+ * lands correctly.
+ */
+async function heatmapCells(
+  db: D1Database,
+  start: string,
+  days: number,
+): Promise<{ cells: HeatmapCell[]; weeks: number }> {
+  const [sy, sm, sd] = start.split("-").map(Number);
+  const base = Date.UTC(sy!, sm! - 1, sd!);
+  const end = new Date(base + (days - 1) * DAY_MS).toISOString().slice(0, 10);
+
+  const { results } = await db
+    .prepare(`SELECT date, COUNT(*) AS attempts FROM attempts WHERE date >= ?1 AND date <= ?2 GROUP BY date`)
+    .bind(start, end)
+    .all<{ date: string; attempts: number }>();
+
+  const byDate = new Map(results.map((r) => [r.date, r.attempts]));
+
+  const cells: HeatmapCell[] = [];
+  for (let i = 0; i < days; i++) {
+    const day = new Date(base + i * DAY_MS);
+    const attempts = byDate.get(day.toISOString().slice(0, 10)) ?? 0;
+    cells.push({
+      col: Math.floor(i / 7),
+      row: (day.getUTCDay() + 6) % 7, // Mon = 0
+      level: Math.min(attempts, 4),
+    });
+  }
+
+  return { cells, weeks: Math.ceil(days / 7) };
+}
+
+export async function heatmapChart(db: D1Database, start: string, days: number): Promise<string> {
+  const { cells, weeks } = await heatmapCells(db, start, days);
+
+  const parts = [svgOpen(HEAT_W, HEAT_H)];
+
+  // Week numbers across the top, Mon/Wed/Fri down the left.
+  for (let w = 0; w < weeks; w++) {
+    parts.push(
+      text(HEAT_GRID_X + w * HEAT_STEP + HEAT_CELL / 2, HEAT_GRID_Y - 7, `W${w + 1}`, {
+        size: 10,
+        fill: MUTED,
+        anchor: "middle",
+      }),
+    );
+  }
+  for (const [row, label] of WEEKDAYS) {
+    parts.push(
+      text(HEAT_GRID_X - 6, HEAT_GRID_Y + row * HEAT_STEP + HEAT_CELL - 4, label, {
+        size: 10,
+        fill: MUTED,
+        anchor: "end",
+      }),
+    );
+  }
+
+  for (const { col, row, level } of cells) {
+    parts.push(
+      `<rect x="${HEAT_GRID_X + col * HEAT_STEP}" y="${HEAT_GRID_Y + row * HEAT_STEP}" width="${HEAT_CELL}" height="${HEAT_CELL}" rx="2" fill="${GREENS[level]}"/>`,
+    );
+  }
+
+  // Less → More ramp fills the space right of the grid.
+  const legendX = HEAT_GRID_X + weeks * HEAT_STEP + 40;
+  const legendY = HEAT_GRID_Y + 3 * HEAT_STEP;
+  parts.push(text(legendX - 6, legendY + HEAT_CELL - 4, "Less", { size: 10, fill: MUTED, anchor: "end" }));
+  GREENS.forEach((color, i) => {
+    parts.push(
+      `<rect x="${legendX + i * HEAT_STEP}" y="${legendY}" width="${HEAT_CELL}" height="${HEAT_CELL}" rx="2" fill="${color}"/>`,
+    );
+  });
+  parts.push(text(legendX + GREENS.length * HEAT_STEP + 3, legendY + HEAT_CELL - 4, "More", { size: 10, fill: MUTED }));
+
+  parts.push("</svg>");
+  return parts.join("");
+}
+
+// The PNG carries no alpha channel, so whatever the rounded card does not cover
+// has to be painted with the colour sitting behind the image: the digest
+// email's GitHub-dark card.
+const EMAIL_CARD = "#0d1117";
+
+// Device-pixel multiplier. The email hands the image the full 552px card
+// width, so 3x lands a little over 2x density on a retina screen; the labels
+// stay 10 CSS px, the size of the SVG's. Flat colour compresses to a few KB
+// either way, so the extra resolution is close to free.
+const HEAT_SCALE = 3;
+
+/**
+ * The same picture as heatmapChart, rastered for email.
+ *
+ * Two deliberate departures from the SVG. The canvas is trimmed to the width
+ * the content actually occupies instead of HEAT_W: the README's SVG sits in a
+ * narrow table cell and scales up, whereas the email gives the image the full
+ * 552px card, and the SVG's slack right margin would otherwise shrink the
+ * grid to nothing. And the bitmap font hangs off a top-left origin rather than
+ * a baseline, so each SVG baseline becomes "baseline minus one cap height".
+ */
+export async function heatmapPng(db: D1Database, start: string, days: number): Promise<Uint8Array> {
+  const { cells, weeks } = await heatmapCells(db, start, days);
+
+  const s = HEAT_SCALE;
+  const gridX = HEAT_GRID_X * s;
+  const gridY = HEAT_GRID_Y * s;
+  const cell = HEAT_CELL * s;
+  const step = HEAT_STEP * s;
+  const capH = 7 * s;
+
+  const legendX = gridX + weeks * step + 40 * s;
+  const legendY = gridY + 3 * step;
+  const moreX = legendX + GREENS.length * step + 3 * s;
+  const width = moreX + textWidth("More", s) + gridX;
+  const height = HEAT_H * s;
+
+  const canvas = new Canvas(width, height, EMAIL_CARD);
+  canvas.rect(0, 0, width, height, BG, 6 * s);
+
+  for (let w = 0; w < weeks; w++) {
+    const label = `W${w + 1}`;
+    const middle = gridX + w * step + cell / 2;
+    canvas.text(middle - textWidth(label, s) / 2, gridY - 7 * s - capH, label, MUTED, s);
+  }
+  for (const [row, label] of WEEKDAYS) {
+    canvas.text(
+      gridX - 6 * s - textWidth(label, s),
+      gridY + row * step + cell - 4 * s - capH,
+      label,
+      MUTED,
+      s,
+    );
+  }
+
+  for (const { col, row, level } of cells) {
+    canvas.rect(gridX + col * step, gridY + row * step, cell, cell, GREENS[level]!, 2 * s);
+  }
+
+  const legendBase = legendY + cell - 4 * s - capH;
+  canvas.text(legendX - 6 * s - textWidth("Less", s), legendBase, "Less", MUTED, s);
+  GREENS.forEach((color, i) => {
+    canvas.rect(legendX + i * step, legendY, cell, cell, color, 2 * s);
+  });
+  canvas.text(moreX, legendBase, "More", MUTED, s);
+
+  return await canvas.encode();
+}
+
+// ── gate badge: shieldcn-style dark pill ─────────────────────────
+
+export async function gateBadge(db: D1Database): Promise<string> {
+  const row = await db
+    .prepare(`SELECT pass_rate FROM gates WHERE pass_rate IS NOT NULL ORDER BY week DESC LIMIT 1`)
+    .first<{ pass_rate: number }>();
+
+  const label = "gate";
+  const value = row ? `${Math.round(row.pass_rate * 100)}%` : "none yet";
+  const valueFill = row ? (row.pass_rate >= 0.7 ? "#4ade80" : RED) : MUTED;
+
+  // shieldcn geometry: height 32, rx 6, one flat zinc-900 pill. Verdana 13px
+  // ≈ 7.5px per char; 12px outer padding, 8px between label and value.
+  const labelW = Math.round(label.length * 7.5);
+  const valueW = Math.round(value.length * 7.5);
+  const total = 12 + labelW + 8 + valueW + 12;
+
+  return (
+    `<svg xmlns="http://www.w3.org/2000/svg" width="${total}" height="32" ` +
+    `viewBox="0 0 ${total} 32" role="img" aria-label="${label}: ${value}" font-family="${FONT}">` +
+    `<rect width="${total}" height="32" rx="6" fill="#18181b"/>` +
+    `<g font-size="13">` +
+    `<text x="12" y="21" fill="${FG}" fill-opacity=".7">${label}</text>` +
+    `<text x="${12 + labelW + 8}" y="21" fill="${valueFill}" font-weight="bold">${value}</text>` +
+    `</g>` +
+    `</svg>`
+  );
+}

apps/api/src/digest.ts

@@ -0,0 +1,195 @@
+/**
+ * The daily digest email — built from D1 + the bundled schedule, sent at
+ * 8 AM ET via the send_email binding. This module owns data only: it turns
+ * D1 rows into a `DigestData` and hands it to email.tsx, which renders both
+ * the HTML and the plain-text alternative from that single tree (React Email
+ * `render`), so the two can never drift.
+ *
+ * Retrieval rules: review and drill rows carry number + difficulty only —
+ * never the topic, never a solution link. `DigestRow` has no field for
+ * either, so the rule holds by construction. The learning day's core list is
+ * the only labeled section. At most REVIEW_CAP (3) reviews per day; drills
+ * fill toward DAILY_CAP (6) only when the due queue is light. Overflow is
+ * levelled, not stacked: sendDigest gives every review past the cap a
+ * concrete future date, ≤ REVIEW_CAP per day, oldest first. Sunday is a
+ * two-line rest note.
+ *
+ * Idempotency: email_log keys sends by ET date — a same-day re-send is a
+ * no-op unless forced; the body itself is deterministic (drill picks are
+ * seeded by the date). Levelling keeps that: it runs once, before the send,
+ * and a re-run finds nothing left over the cap.
+ */
+import {
+  CAMPAIGN_DAYS,
+  type ProblemRow,
+  REVIEW_CAP,
+  SCHEDULE,
+  addDays,
+  campaignDay,
+  campaignWeek,
+  dueReviews,
+  isoWeek,
+  levelReviews,
+  pickDrills,
+  prettyDate,
+  streak,
+  weekdayOf,
+} from "./srs.ts";
+import { type DigestData, type DigestRow, renderDigest } from "./email.tsx";
+import { signLink } from "./links.ts";
+
+/** Total daily ceiling: drills only fill what the whole due queue leaves. */
+const DAILY_CAP = 6;
+
+export interface Digest {
+  subject: string;
+  html: string;
+  text: string;
+}
+
+/** A problem row plus its signed one-tap pass/fail URLs. */
+async function tapRow(env: Env, p: ProblemRow, date: string, staged: boolean): Promise<DigestRow> {
+  const [pass, fail] = await Promise.all([
+    signLink(env.LINK_KEY, p.lc_number, "pass", date),
+    signLink(env.LINK_KEY, p.lc_number, "fail", date),
+  ]);
+  const base = `${env.PUBLIC_URL}/log?p=${p.lc_number}&d=${date}&r=`;
+  return {
+    lc: p.lc_number,
+    difficulty: p.difficulty,
+    ...(staged ? { stage: p.stage } : {}),
+    passUrl: `${base}pass&sig=${pass}`,
+    failUrl: `${base}fail&sig=${fail}`,
+  };
+}
+
+/** Everything the email needs, read straight out of D1 and the schedule. */
+export async function collectDigest(env: Env, date: string): Promise<DigestData> {
+  const db = env.DB;
+  const rest = weekdayOf(date) === 0;
+
+  // Reviews surface at most REVIEW_CAP; drills fill toward DAILY_CAP only
+  // when the WHOLE due queue is light — a heavy backlog must never trade
+  // review slots for brand-new drill problems.
+  const due = rest ? [] : await dueReviews(db, date);
+  const capped = due.slice(0, REVIEW_CAP);
+  const drills = rest ? [] : await pickDrills(db, date, DAILY_CAP - Math.min(due.length, DAILY_CAP));
+
+  const data: DigestData = {
+    day: prettyDate(date),
+    progress: `Day ${campaignDay(date)} of ${CAMPAIGN_DAYS} · Week ${campaignWeek(date)}`,
+    streak: await streak(db, date),
+    rest,
+    reviews: await Promise.all(capped.map((p) => tapRow(env, p, date, true))),
+    carried: due.length - capped.length,
+    drills: await Promise.all(drills.map((p) => tapRow(env, p, date, false))),
+    yesterday: { total: 0, failed: [] },
+    heatmapUrl: `${env.PUBLIC_URL}/chart/heatmap.png`,
+    progressUrl: `${env.DOCS_URL}/progress`,
+  };
+
+  // New topic — the only section allowed to name problems and link them.
+  const topicIssue = rest ? undefined : SCHEDULE[date];
+  if (topicIssue !== undefined) {
+    const topic = await db
+      .prepare("SELECT name FROM topics WHERE issue = ?")
+      .bind(topicIssue)
+      .first<{ name: string }>();
+    const { results: core } = await db
+      .prepare(
+        "SELECT * FROM problems WHERE topic_issue = ? AND set_label = 'core' ORDER BY lc_number",
+      )
+      .bind(topicIssue)
+      .all<ProblemRow>();
+    data.topic = {
+      name: topic?.name ?? `#${topicIssue}`,
+      core: core.map((p) => ({
+        lc: p.lc_number,
+        difficulty: p.difficulty,
+        url: `https://github.com/${env.REPO}/issues/${p.issue}`,
+        solved: p.stage !== "new",
+      })),
+    };
+  }
+
+  // Saturday: the review issue already exists (created at midnight ET).
+  if (weekdayOf(date) === 6) {
+    const gate = await db
+      .prepare("SELECT issue FROM gates WHERE week = ? AND issue IS NOT NULL")
+      .bind(isoWeek(date))
+      .first<{ issue: number }>();
+    if (gate) {
+      data.gate = {
+        week: campaignWeek(date),
+        url: `https://github.com/${env.REPO}/issues/${gate.issue}`,
+      };
+    }
+  }
+
+  // Footer: yesterday's log summarised (failures named — they are the
+  // actionable part) and the gate rate to date.
+  const { results: logged } = await db
+    .prepare("SELECT lc_number, result FROM attempts WHERE date = ? ORDER BY id")
+    .bind(addDays(date, -1))
+    .all<{ lc_number: number; result: string }>();
+  data.yesterday = {
+    total: logged.length,
+    failed: logged.filter((a) => a.result !== "pass").map((a) => a.lc_number),
+  };
+  const lastGate = await db
+    .prepare("SELECT pass_rate FROM gates WHERE pass_rate IS NOT NULL ORDER BY week DESC LIMIT 1")
+    .first<{ pass_rate: number }>();
+  if (lastGate) data.gateRate = lastGate.pass_rate;
+
+  return data;
+}
+
+export async function buildDigest(env: Env, date: string): Promise<Digest> {
+  const data = await collectDigest(env, date);
+  const { html, text } = await renderDigest(data);
+  return {
+    subject: `(Day ${campaignDay(date)}/${CAMPAIGN_DAYS}) LeetCode Daily Digest`,
+    html,
+    text,
+  };
+}
+
+export interface SendReport {
+  sent: boolean;
+  reason: string;
+  digest: Digest;
+}
+
+/** Send today's digest exactly once per ET date (unless forced). */
+export async function sendDigest(
+  env: Env,
+  date: string,
+  opts: { force?: boolean; dry?: boolean } = {},
+): Promise<SendReport> {
+  const digest = await buildDigest(env, date);
+  if (opts.dry) return { sent: false, reason: "dry run", digest };
+
+  const already = await env.DB.prepare("SELECT sent_at FROM email_log WHERE date = ?")
+    .bind(date)
+    .first();
+  if (already && !opts.force) return { sent: false, reason: "already sent today", digest };
+
+  // The digest above was built against the un-levelled queue (so `carried`
+  // reports the real backlog); now give everything past the cap its future
+  // date. Dry runs never reach here, and a forced re-send finds nothing left.
+  await levelReviews(env.DB, date);
+
+  await env.EMAIL.send({
+    to: env.TO_EMAIL,
+    from: { email: env.FROM_EMAIL, name: "SRS" },
+    subject: digest.subject,
+    html: digest.html,
+    text: digest.text,
+  });
+  await env.DB.prepare(
+    "INSERT INTO email_log (date, sent_at) VALUES (?, ?) ON CONFLICT(date) DO UPDATE SET sent_at = excluded.sent_at",
+  )
+    .bind(date, new Date().toISOString())
+    .run();
+  return { sent: true, reason: already ? "forced re-send" : "sent", digest };
+}

apps/api/src/email.tsx

@@ -0,0 +1,495 @@
+/**
+ * Presentation layer for the daily digest — React Email components rendered
+ * to HTML (and to the plain-text alternative) inside the Worker.
+ *
+ * This file owns *only* layout and wording. Every number, URL and signature
+ * is computed in digest.ts and handed over as `DigestData`, so the retrieval
+ * rules (review/drill rows carry number + difficulty only — never the topic,
+ * never a solution link) are enforced by what the data model can express:
+ * `DigestRow` has no title and no issue field.
+ *
+ * Dark by design. The theme is hard-coded rather than left to the client's
+ * dark-mode heuristics: `color-scheme: dark` tells Apple Mail and Outlook not
+ * to re-invert it, and the canvas colour is painted by a full-width <Section>
+ * table because Gmail drops styles on <body>.
+ *
+ * Other email constraints that shape the markup: inline styles only (clients
+ * strip <style>), real <table> layout for the problem lists so Outlook and
+ * Gmail agree on column alignment, React Email's <Button> for the one-tap
+ * links (it emits the MSO padding conditionals a bare <a> lacks), and the
+ * heatmap arrives as PNG — every major client refuses remote SVG.
+ */
+import {
+  Body,
+  Button,
+  Container,
+  Head,
+  Heading,
+  Hr,
+  Html,
+  Img,
+  Link,
+  Preview,
+  Section,
+  Text,
+} from "@react-email/components";
+import { render } from "@react-email/render";
+import type { CSSProperties } from "react";
+import { REVIEW_CAP } from "./srs.ts";
+
+// ── palette (GitHub dark) ────────────────────────────────────────
+const CANVAS = "#010409";
+const CARD = "#0d1117";
+const BORDER = "#30363d";
+const RULE = "#21262d";
+const INK = "#e6edf3";
+const MUTED = "#8b949e";
+const LINK = "#58a6ff";
+const GREEN = "#3fb950";
+const AMBER = "#d29922";
+const RED = "#f85149";
+const PASS_BG = "#238636";
+const FAIL_BG = "#da3633";
+const GATE_BG = "#1f6feb";
+
+const FONT = '-apple-system,BlinkMacSystemFont,"Segoe UI",Helvetica,Arial,sans-serif';
+
+// Outlook's Word engine is unreliable with text-transform, so display labels
+// are table lookups rather than a CSS trick.
+const DIFFICULTY: Record<string, { label: string; color: string }> = {
+  easy: { label: "Easy", color: GREEN },
+  medium: { label: "Medium", color: AMBER },
+  hard: { label: "Hard", color: RED },
+};
+
+// The SRS ladder in plain English — stage "+3" means the last clean look was
+// three days back, so that is what the reader is told.
+const LAST_SEEN: Record<string, string> = {
+  new: "first look",
+  "+3": "3 days ago",
+  "+7": "7 days ago",
+  retired: "retired",
+};
+
+// ── data model ───────────────────────────────────────────────────
+
+/** A review or drill line: number + difficulty + one-tap links, nothing else. */
+export interface DigestRow {
+  lc: number;
+  difficulty: string;
+  /** Reviews only — drills are unstaged by design. */
+  stage?: string;
+  passUrl: string;
+  failUrl: string;
+}
+
+/** A learning-day core problem — the only rows allowed a link. */
+export interface TopicRow {
+  lc: number;
+  difficulty: string;
+  url: string;
+  solved: boolean;
+}
+
+export interface DigestData {
+  /** "Tuesday, August 25" */
+  day: string;
+  /** "Day 9 of 56 · Week 2" */
+  progress: string;
+  streak: number;
+  rest: boolean;
+  topic?: { name: string; core: TopicRow[] };
+  reviews: DigestRow[];
+  /** Reviews past the daily cap; levelled onto the coming days. */
+  carried: number;
+  drills: DigestRow[];
+  gate?: { week: number; url: string };
+  yesterday: { total: number; failed: number[] };
+  gateRate?: number;
+  /** PNG, not SVG — email clients refuse the latter. */
+  heatmapUrl: string;
+  progressUrl: string;
+}
+
+// ── styles ───────────────────────────────────────────────────────
+
+const page: CSSProperties = { backgroundColor: CANVAS, padding: "28px 0" };
+const card: CSSProperties = {
+  backgroundColor: CARD,
+  border: `1px solid ${BORDER}`,
+  borderRadius: "10px",
+  margin: "0 auto",
+  maxWidth: "600px",
+  padding: "28px 24px",
+};
+const h1: CSSProperties = { color: INK, fontSize: "20px", fontWeight: 600, lineHeight: "26px", margin: 0 };
+const label: CSSProperties = {
+  color: MUTED,
+  fontSize: "11px",
+  fontWeight: 700,
+  letterSpacing: "0.8px",
+  margin: "0 0 10px",
+  textTransform: "uppercase",
+};
+const hint: CSSProperties = { color: MUTED, fontSize: "13px", lineHeight: "19px", margin: "0 0 12px" };
+const th: CSSProperties = {
+  borderBottom: `1px solid ${BORDER}`,
+  color: MUTED,
+  fontSize: "12px",
+  fontWeight: 400,
+  padding: "0 0 8px",
+  textAlign: "left",
+};
+const td: CSSProperties = {
+  borderBottom: `1px solid ${RULE}`,
+  color: INK,
+  fontSize: "14px",
+  padding: "12px 0",
+  textAlign: "left",
+};
+const tap: CSSProperties = {
+  borderRadius: "6px",
+  color: "#ffffff",
+  display: "inline-block",
+  fontSize: "12px",
+  fontWeight: 600,
+  lineHeight: "12px",
+  padding: "9px 15px",
+  textDecoration: "none",
+};
+
+/**
+ * A problem list as a real table. The "last seen" column appears only for
+ * reviews, which keeps drills at three columns on a phone.
+ */
+function ProblemTable({ rows, staged }: { rows: DigestRow[]; staged: boolean }) {
+  return (
+    <table
+      width="100%"
+      border={0}
+      cellPadding={0}
+      cellSpacing={0}
+      style={{ borderCollapse: "collapse", width: "100%" }}
+    >
+      <thead>
+        <tr>
+          <th style={th}>Problem</th>
+          <th style={th}>Level</th>
+          {staged ? <th style={th}>Last seen</th> : null}
+          <th style={{ ...th, textAlign: "right" }}>How did it go?</th>
+        </tr>
+      </thead>
+      <tbody>
+        {rows.map((row) => {
+          const difficulty = DIFFICULTY[row.difficulty];
+          return (
+            <tr key={row.lc}>
+              <td style={{ ...td, fontWeight: 600 }}>LC {row.lc}</td>
+              <td style={{ ...td, color: difficulty?.color ?? INK }}>{difficulty?.label ?? row.difficulty}</td>
+              {staged ? (
+                <td style={{ ...td, color: MUTED }}>{(row.stage && LAST_SEEN[row.stage]) ?? row.stage}</td>
+              ) : null}
+              <td style={{ ...td, textAlign: "right", whiteSpace: "nowrap" }}>
+                <Button href={row.passUrl} style={{ ...tap, backgroundColor: PASS_BG }}>
+                  got it
+                </Button>
+                <Button href={row.failUrl} style={{ ...tap, backgroundColor: FAIL_BG, marginLeft: "6px" }}>
+                  missed it
+                </Button>
+              </td>
+            </tr>
+          );
+        })}
+      </tbody>
+    </table>
+  );
+}
+
+/**
+ * The learning day's core list as the same table shape as reviews — this is
+ * the one section allowed to link problems, and ticks replace tap buttons.
+ */
+function TopicTable({ rows }: { rows: TopicRow[] }) {
+  return (
+    <table
+      width="100%"
+      border={0}
+      cellPadding={0}
+      cellSpacing={0}
+      style={{ borderCollapse: "collapse", width: "100%" }}
+    >
+      <thead>
+        <tr>
+          <th style={th}>Problem</th>
+          <th style={th}>Level</th>
+          <th style={{ ...th, textAlign: "right" }}>Solved</th>
+        </tr>
+      </thead>
+      <tbody>
+        {rows.map((p) => {
+          const difficulty = DIFFICULTY[p.difficulty];
+          return (
+            <tr key={p.lc}>
+              <td style={{ ...td, fontWeight: 600 }}>
+                <Link href={p.url} style={{ color: p.solved ? MUTED : LINK, textDecoration: "none" }}>
+                  LC {p.lc}
+                </Link>
+              </td>
+              <td style={{ ...td, color: difficulty?.color ?? INK }}>{difficulty?.label ?? p.difficulty}</td>
+              <td style={{ ...td, color: p.solved ? GREEN : MUTED, textAlign: "right" }}>
+                {p.solved ? "✓" : "—"}
+              </td>
+            </tr>
+          );
+        })}
+      </tbody>
+    </table>
+  );
+}
+
+function Block({
+  title,
+  note,
+  rows,
+  staged,
+  empty,
+}: {
+  title: string;
+  note: string;
+  rows: DigestRow[];
+  staged: boolean;
+  empty: string;
+}) {
+  return (
+    <Section style={{ marginTop: "28px" }}>
+      <Text style={label}>{title}</Text>
+      {rows.length === 0 ? (
+        <Text style={{ ...hint, margin: 0 }}>{empty}</Text>
+      ) : (
+        <>
+          <Text style={hint}>{note}</Text>
+          <ProblemTable rows={rows} staged={staged} />
+        </>
+      )}
+    </Section>
+  );
+}
+
+// ── email ────────────────────────────────────────────────────────
+
+function DigestEmail({ data }: { data: DigestData }) {
+  const load = data.reviews.length + data.drills.length;
+  const preview = data.rest
+    ? "Rest day — nothing to do but rest."
+    : `${load} to work through today${data.topic ? `, starting with ${data.topic.name}` : ""}.`;
+
+  const passed = data.yesterday.total - data.yesterday.failed.length;
+  const missed = data.yesterday.failed.map((lc) => `LC ${lc}`).join(", ");
+
+  return (
+    <Html lang="en" dir="ltr">
+      <Head>
+        <meta name="color-scheme" content="dark" />
+        <meta name="supported-color-schemes" content="dark" />
+      </Head>
+      <Preview>{preview}</Preview>
+      <Body style={{ backgroundColor: CANVAS, fontFamily: FONT, margin: 0, padding: 0 }}>
+        <Section style={page}>
+          <Container style={card}>
+            <Heading as="h1" style={h1}>
+              {data.rest ? `Take the day off — ${data.day}` : `Good morning. It's ${data.day}.`}
+            </Heading>
+            <Text style={{ color: MUTED, fontSize: "13px", margin: "8px 0 0" }}>
+              {data.progress}
+              {data.rest
+                ? " · the ladder never books a Sunday, so nothing is due — anything still open waits for Monday."
+                : data.streak === 0
+                  ? " · no streak going yet — today is a good day to start one."
+                  : ` · 🔥 ${data.streak} day${data.streak === 1 ? "" : "s"} in a row, keep it going.`}
+            </Text>
+
+            {data.rest ? null : (
+              <>
+                {data.topic ? (
+                  <Section style={{ marginTop: "28px" }}>
+                    <Text style={label}>Something new today</Text>
+                    <Text style={{ color: INK, fontSize: "16px", fontWeight: 600, margin: "0 0 8px" }}>
+                      {data.topic.name}
+                    </Text>
+                    <Text style={hint}>Work through these from the top — a tick means it is already solved.</Text>
+                    <TopicTable rows={data.topic.core} />
+                  </Section>
+                ) : null}
+
+                <Block
+                  title="Time to see these again"
+                  note="Solve each one from scratch, and resist opening your old answer first."
+                  rows={data.reviews}
+                  staged
+                  empty="Nothing is due for review today."
+                />
+                {data.carried > 0 ? (
+                  <Text style={{ ...hint, margin: "12px 0 0" }}>
+                    {data.carried} more {data.carried === 1 ? "is" : "are"} waiting — spread over the
+                    coming days, oldest first, never more than {REVIEW_CAP} a day.
+                  </Text>
+                ) : null}
+
+                <Block
+                  title="Cold start, no hints"
+                  note="You are not told the topic. Say the pattern out loud before you write a line."
+                  rows={data.drills}
+                  staged={false}
+                  empty="No cold starts today."
+                />
+
+                {data.gate ? (
+                  <Section style={{ marginTop: "28px" }}>
+                    <Text style={label}>Weekly checkpoint</Text>
+                    <Text style={hint}>Timed and blind. Close the issue when you are done.</Text>
+                    <Button
+                      href={data.gate.url}
+                      style={{ ...tap, backgroundColor: GATE_BG, fontSize: "13px", padding: "11px 18px" }}
+                    >
+                      Open week {data.gate.week} review
+                    </Button>
+                  </Section>
+                ) : null}
+              </>
+            )}
+
+            {/* Eight weeks at a glance — the one thing here that is pure
+                encouragement rather than instruction. */}
+            <Section style={{ marginTop: "28px" }}>
+              <Text style={label}>Every day you showed up</Text>
+              <Link href={data.progressUrl}>
+                <Img
+                  src={data.heatmapUrl}
+                  alt="Attempt heatmap across the eight-week campaign"
+                  width="552"
+                  style={{ border: `1px solid ${BORDER}`, borderRadius: "8px", display: "block", width: "100%" }}
+                />
+              </Link>
+            </Section>
+
+            <Hr style={{ borderColor: BORDER, margin: "28px 0 18px" }} />
+            <Text style={{ color: MUTED, fontSize: "12px", lineHeight: "19px", margin: 0 }}>
+              {data.yesterday.total === 0
+                ? "You did not log anything yesterday."
+                : `Yesterday you logged ${data.yesterday.total} and got ${passed} of them` +
+                  (missed ? `; ${missed} got away.` : ".")}{" "}
+              {data.gateRate === undefined
+                ? "No checkpoints scored yet."
+                : `Checkpoints are running at ${Math.round(data.gateRate * 100)}%.`}{" "}
+              <Link href={data.progressUrl} style={{ color: LINK, textDecoration: "none" }}>
+                See the full picture →
+              </Link>
+            </Text>
+          </Container>
+        </Section>
+      </Body>
+    </Html>
+  );
+}
+
+/**
+ * The plain-text alternative. React Email's `plainText` mode flattens the
+ * problem tables into one unreadable run, so text gets its own writer — fed
+ * by the same `DigestData`, so the numbers and links can never disagree with
+ * the HTML even though the wording lives in two places.
+ */
+function plainDigest(data: DigestData): string {
+  const out: string[] = [
+    data.rest ? `Take the day off — ${data.day}` : `Good morning. It's ${data.day}.`,
+    data.progress +
+      (data.rest
+        ? " · the ladder never books a Sunday, so nothing is due — anything still open waits for Monday."
+        : data.streak === 0
+          ? " · no streak going yet — today is a good day to start one."
+          : ` · ${data.streak} day${data.streak === 1 ? "" : "s"} in a row, keep it going.`),
+  ];
+
+  if (!data.rest) {
+    if (data.topic) {
+      out.push(
+        "",
+        `SOMETHING NEW TODAY — ${data.topic.name}`,
+        "Work through these from the top; the ones marked done you have already solved.",
+        ...data.topic.core.map(
+          (p) =>
+            `  LC ${p.lc} — ${DIFFICULTY[p.difficulty]?.label ?? p.difficulty}${p.solved ? " (done)" : ""} — ${p.url}`,
+        ),
+      );
+    }
+
+    const lists: [string, string, DigestRow[], string][] = [
+      [
+        "TIME TO SEE THESE AGAIN",
+        "Solve each one from scratch, and resist opening your old answer first.",
+        data.reviews,
+        "Nothing is due for review today.",
+      ],
+      [
+        "COLD START, NO HINTS",
+        "You are not told the topic. Say the pattern out loud before you write a line.",
+        data.drills,
+        "No cold starts today.",
+      ],
+    ];
+    for (const [title, note, rows, empty] of lists) {
+      out.push("", title);
+      if (rows.length === 0) {
+        out.push(empty);
+        continue;
+      }
+      out.push(note);
+      for (const row of rows) {
+        const seen = row.stage ? `, last seen ${LAST_SEEN[row.stage] ?? row.stage}` : "";
+        out.push(
+          `  LC ${row.lc} — ${DIFFICULTY[row.difficulty]?.label ?? row.difficulty}${seen}`,
+          `    got it:    ${row.passUrl}`,
+          `    missed it: ${row.failUrl}`,
+        );
+      }
+    }
+    if (data.carried > 0) {
+      out.push(
+        "",
+        `${data.carried} more ${data.carried === 1 ? "is" : "are"} waiting — spread over the coming days, oldest first, never more than ${REVIEW_CAP} a day.`,
+      );
+    }
+    if (data.gate) {
+      out.push(
+        "",
+        "WEEKLY CHECKPOINT",
+        "Timed and blind. Close the issue when you are done.",
+        `  Week ${data.gate.week} review — ${data.gate.url}`,
+      );
+    }
+  }
+
+  const passed = data.yesterday.total - data.yesterday.failed.length;
+  const missed = data.yesterday.failed.map((lc) => `LC ${lc}`).join(", ");
+  out.push(
+    "",
+    "─".repeat(48),
+    data.yesterday.total === 0
+      ? "You did not log anything yesterday."
+      : `Yesterday you logged ${data.yesterday.total} and got ${passed} of them` +
+        (missed ? `; ${missed} got away.` : "."),
+    data.gateRate === undefined
+      ? "No checkpoints scored yet."
+      : `Checkpoints are running at ${Math.round(data.gateRate * 100)}%.`,
+    `Every day you showed up: ${data.heatmapUrl}`,
+    `See the full picture: ${data.progressUrl}`,
+  );
+  return out.join("\n");
+}
+
+/**
+ * The HTML body plus its plain-text alternative — always sent as a pair so a
+ * client that refuses HTML still gets the same problems and the same links.
+ */
+export async function renderDigest(data: DigestData): Promise<{ html: string; text: string }> {
+  return { html: await render(<DigestEmail data={data} />), text: plainDigest(data) };
+}

apps/api/src/gate.ts

@@ -0,0 +1,270 @@
+/**
+ * Saturday review issue: created at midnight ET so the 8 AM digest can link
+ * to it; scored when the issue closes (webhook).
+ *
+ * The gate half is a topic-blind quiz — one unsolved optional problem per
+ * topic scheduled this week (which is why daily drills only draw from
+ * earlier weeks). The recap half is generated data, deliberately NOT a task
+ * list: the week's solved problems are already scheduled by the +3/+7 ladder,
+ * and re-assigning them on Saturday would be massed practice.
+ *
+ * Label is `review` (the retired Actions system owned `gate`). Creation is
+ * idempotent by week; deferred Hards never appear.
+ */
+import {
+  type ProblemRow,
+  type Result,
+  SCHEDULE,
+  campaignWeek,
+  isoWeek,
+  rng,
+  sample,
+  streak,
+  weekMonday,
+} from "./srs.ts";
+import type { GitHub } from "./github.ts";
+
+const PASS_TARGET = 0.7;
+const REVIEW_LABEL = "review";
+const MAX_GATE_PROBLEMS = 5;
+
+function capitalized(difficulty: string): string {
+  return difficulty[0]!.toUpperCase() + difficulty.slice(1);
+}
+
+/** Unsolved problems of a topic in a set — never attempted, never drilled. */
+async function freshPool(db: D1Database, topic: number, set: string): Promise<ProblemRow[]> {
+  const { results } = await db
+    .prepare(
+      `SELECT * FROM problems
+       WHERE topic_issue = ?1 AND set_label = ?2 AND stage = 'new' AND defer_until IS NULL
+         AND lc_number NOT IN (SELECT lc_number FROM drill_pool_used)
+         AND NOT EXISTS (SELECT 1 FROM attempts a WHERE a.lc_number = problems.lc_number)
+       ORDER BY lc_number`,
+    )
+    .bind(topic, set)
+    .all<ProblemRow>();
+  return results;
+}
+
+export interface CreateReport {
+  created: boolean;
+  issue?: number;
+  reason: string;
+}
+
+export async function createReviewIssue(env: Env, gh: GitHub, date: string): Promise<CreateReport> {
+  const db = env.DB;
+  const week = campaignWeek(date);
+  const iso = isoWeek(date);
+  const title = `Review — Week ${week}`;
+
+  const existing = await db
+    .prepare("SELECT issue FROM gates WHERE week = ? AND issue IS NOT NULL")
+    .bind(iso)
+    .first<{ issue: number }>();
+  if (existing) {
+    return { created: false, issue: existing.issue, reason: `${title} exists (#${existing.issue})` };
+  }
+
+  // Topics scheduled this week; review-only weeks (Sep 23 onward) sample
+  // across every topic that still has unsolved optional problems.
+  let topics = Object.entries(SCHEDULE)
+    .filter(([d]) => campaignWeek(d) === week)
+    .sort(([a], [b]) => (a < b ? -1 : 1))
+    .map(([, t]) => t);
+  if (topics.length === 0) {
+    const { results } = await db
+      .prepare(
+        `SELECT DISTINCT topic_issue AS t FROM problems
+         WHERE set_label = 'optional' AND stage = 'new'
+           AND lc_number NOT IN (SELECT lc_number FROM drill_pool_used)
+         ORDER BY topic_issue`,
+      )
+      .all<{ t: number }>();
+    topics = sample(results.map((r) => r.t), MAX_GATE_PROBLEMS, rng(`gate-topics-${iso}`)).sort(
+      (a, b) => a - b,
+    );
+  }
+  if (topics.length === 0) return { created: false, reason: "no topics with unsolved pools" };
+
+  const random = rng(`gate-${iso}`);
+  const picks: { p: ProblemRow; fallback: boolean }[] = [];
+  for (const topic of topics) {
+    const optional = await freshPool(db, topic, "optional");
+    if (optional.length > 0) {
+      picks.push({ p: sample(optional, 1, random)[0]!, fallback: false });
+      continue;
+    }
+    const core = await freshPool(db, topic, "core");
+    if (core.length > 0) picks.push({ p: sample(core, 1, random)[0]!, fallback: true });
+    // Both pools exhausted: the topic is fully solved; nothing to quiz.
+  }
+  if (picks.length === 0) return { created: false, reason: `every pool for week ${week} is exhausted` };
+
+  // Recap: everything attempted this week, plus streak — data, not tasks.
+  const monday = weekMonday(date);
+  const { results: attempts } = await db
+    .prepare(
+      `SELECT a.lc_number, a.date, a.kind, a.result FROM attempts a
+       WHERE a.date >= ? AND a.date < ? AND a.source != 'import'
+       ORDER BY a.date, a.id`,
+    )
+    .bind(monday, date)
+    .all<{ lc_number: number; date: string; kind: string; result: string }>();
+  const reviewsDone = attempts.filter((a) => a.kind === "review" && a.result === "pass").length;
+  const currentStreak = await streak(db, date);
+
+  const gateLines = picks.map(
+    ({ p, fallback }) =>
+      `- LC ${p.lc_number} — ${capitalized(p.difficulty)}${fallback ? " *(core fallback — optional pool exhausted)*" : ""}`,
+  );
+  const recapLines = attempts.length
+    ? attempts.map(
+        (a) => `| LC ${a.lc_number} | ${a.kind} | ${a.result === "pass" ? "✅" : "❌"} | ${a.date} |`,
+      )
+    : ["| — | no attempts logged this week | | |"];
+  const body = [
+    "## Gate — blind set",
+    `**Target: ${Math.round(PASS_TARGET * 100)}% first-attempt pass rate. Timed. Narrate out loud.**`,
+    "No topics given. Log with `/done <n> pass|fail`, then close this issue.",
+    ...gateLines,
+    "",
+    `## Week ${week} recap`,
+    `${attempts.length} attempts · ${reviewsDone} reviews passed · streak ${currentStreak}`,
+    "",
+    "| Problem | Kind | Result | Day |",
+    "| --- | --- | --- | --- |",
+    ...recapLines,
+  ].join("\n");
+
+  // Ensure the `review` label exists (422 = already there).
+  try {
+    await gh.rest(`/repos/${gh.repo}/labels`, {
+      method: "POST",
+      body: JSON.stringify({
+        name: REVIEW_LABEL,
+        color: "5319e7",
+        description: "Weekly blind gate + recap",
+      }),
+    });
+  } catch (err) {
+    if (!String(err).includes("422")) throw err;
+  }
+
+  const milestone = await db
+    .prepare("SELECT milestone FROM topics WHERE issue = ? AND milestone IS NOT NULL")
+    .bind(topics[0])
+    .first<{ milestone: number }>();
+
+  const created = (await gh.rest(`/repos/${gh.repo}/issues`, {
+    method: "POST",
+    body: JSON.stringify({
+      title,
+      body,
+      labels: [REVIEW_LABEL],
+      milestone: milestone?.milestone,
+    }),
+  })) as { number: number };
+
+  await db
+    .prepare(
+      `INSERT INTO gates (week, issue, problems) VALUES (?1, ?2, ?3)
+       ON CONFLICT(week) DO UPDATE SET issue = ?2, problems = ?3`,
+    )
+    .bind(iso, created.number, JSON.stringify(picks.map(({ p }) => p.lc_number)))
+    .run();
+
+  // Last week's boosted topics had their remedial week — clear the flags.
+  await db.prepare("UPDATE topics SET boost = 0 WHERE boost = 1").run();
+
+  return { created: true, issue: created.number, reason: `created #${created.number}` };
+}
+
+export interface ScoreReport {
+  scored: boolean;
+  rate?: number;
+  reason: string;
+}
+
+/**
+ * Grade a closed review issue: first-attempt pass rate over its gate set.
+ * Unlogged problems count as failures — skipping a gate problem is not a
+ * pass. Boosts topics that failed a gate problem or reached 2 drill misses;
+ * miss counters reset once consumed. The badge/charts read D1 live, so
+ * "refreshing the badge" is this row update.
+ */
+export async function scoreReview(
+  env: Env,
+  gh: GitHub,
+  issueNumber: number,
+  date: string,
+): Promise<ScoreReport> {
+  const db = env.DB;
+  const gate = await db
+    .prepare("SELECT week, problems, pass_rate FROM gates WHERE issue = ?")
+    .bind(issueNumber)
+    .first<{ week: number; problems: string; pass_rate: number | null }>();
+  if (!gate) return { scored: false, reason: `issue #${issueNumber} has no gate record` };
+  if (gate.pass_rate !== null) return { scored: false, rate: gate.pass_rate, reason: "already scored" };
+
+  const lcs: number[] = JSON.parse(gate.problems);
+  const results: Record<number, Result | undefined> = {};
+  for (const lc of lcs) {
+    const row = await db
+      .prepare("SELECT result FROM attempts WHERE lc_number = ? AND kind = 'gate' ORDER BY id LIMIT 1")
+      .bind(lc)
+      .first<{ result: Result }>();
+    results[lc] = row?.result;
+  }
+  const passes = lcs.filter((lc) => results[lc] === "pass");
+  const unlogged = lcs.filter((lc) => results[lc] === undefined);
+  const rate = passes.length / lcs.length;
+
+  await db
+    .prepare("UPDATE gates SET pass_rate = ?, closed_on = ? WHERE issue = ?")
+    .bind(rate, date, issueNumber)
+    .run();
+
+  const boosted: number[] = [];
+  for (const lc of lcs) {
+    if (results[lc] !== "fail") continue;
+    const p = await db
+      .prepare("SELECT topic_issue FROM problems WHERE lc_number = ?")
+      .bind(lc)
+      .first<{ topic_issue: number }>();
+    if (p) boosted.push(p.topic_issue);
+  }
+  const { results: missed } = await db
+    .prepare("SELECT issue FROM topics WHERE misses >= 2")
+    .all<{ issue: number }>();
+  boosted.push(...missed.map((m) => m.issue));
+  const boostSet = [...new Set(boosted)];
+  if (boostSet.length) {
+    await db.batch(
+      boostSet.map((t) => db.prepare("UPDATE topics SET boost = 1, misses = 0 WHERE issue = ?").bind(t)),
+    );
+  }
+
+  const passed = rate >= PASS_TARGET;
+  const summary = [
+    `## Gate — Week ${gate.week}: **${Math.round(rate * 100)}%** first-attempt (target ${Math.round(PASS_TARGET * 100)}%) — ${passed ? "✅ pass" : "❌ fail"}`,
+    "",
+    ...lcs.map((lc) => {
+      const r = results[lc];
+      return `- LC ${lc} — ${r === "pass" ? "✅ pass" : r === "fail" ? "❌ fail" : "⬜ not logged (counted as fail)"}`;
+    }),
+    "",
+    unlogged.length ? `${unlogged.length} problem(s) were never logged.` : "",
+    boostSet.length
+      ? `Boosted topics for next week's drills: ${boostSet.map((t) => `#${t}`).join(", ")}.`
+      : "No topics boosted.",
+    "",
+    `Live charts: ${env.PUBLIC_URL}/chart/progress.svg · progress: ${env.DOCS_URL}/progress`,
+  ]
+    .filter((l) => l !== "")
+    .join("\n");
+  await gh.comment(issueNumber, summary);
+
+  return { scored: true, rate, reason: `scored ${Math.round(rate * 100)}%` };
+}

apps/api/src/github.ts

@@ -0,0 +1,94 @@
+/**
+ * GitHub REST + GraphQL client on GH_PAT. Hand-rolled fetch, no SDK —
+ * matching the repo's dependency-free automation rule.
+ */
+
+export interface GitHub {
+  repo: string;
+  rest(path: string, init?: RequestInit): Promise<unknown>;
+  graphql(query: string, variables?: Record<string, unknown>): Promise<unknown>;
+  list(path: string): AsyncGenerator<unknown, void, void>;
+  /** Comment first, then close — a failed close still leaves a visible note. */
+  closeIssue(issue: number, comment: string): Promise<void>;
+  comment(issue: number, body: string): Promise<void>;
+  react(commentId: number, content: string): Promise<void>;
+}
+
+const PER_PAGE = 100;
+
+export function github(token: string, repo: string): GitHub {
+  async function rest(path: string, init: RequestInit = {}): Promise<unknown> {
+    const res = await fetch(`https://api.github.com${path}`, {
+      ...init,
+      headers: {
+        accept: "application/vnd.github+json",
+        authorization: `Bearer ${token}`,
+        "user-agent": "srs-api",
+        "x-github-api-version": "2022-11-28",
+        ...(init.body ? { "content-type": "application/json" } : {}),
+      },
+    });
+    if (!res.ok) {
+      throw new Error(`${init.method ?? "GET"} ${path} -> ${res.status} ${await res.text()}`);
+    }
+    return res.status === 204 ? null : res.json();
+  }
+
+  async function graphql(
+    query: string,
+    variables: Record<string, unknown> = {},
+  ): Promise<unknown> {
+    const res = await fetch("https://api.github.com/graphql", {
+      method: "POST",
+      headers: {
+        authorization: `Bearer ${token}`,
+        "content-type": "application/json",
+        "user-agent": "srs-api",
+      },
+      body: JSON.stringify({ query, variables }),
+    });
+    if (!res.ok) throw new Error(`graphql -> ${res.status} ${await res.text()}`);
+    const payload = (await res.json()) as { data?: unknown; errors?: { message: string }[] };
+    if (payload.errors?.length) throw new Error(payload.errors.map((e) => e.message).join("; "));
+    return payload.data;
+  }
+
+  async function* list(path: string): AsyncGenerator<unknown, void, void> {
+    const sep = path.includes("?") ? "&" : "?";
+    for (let page = 1; ; page++) {
+      const batch = await rest(`${path}${sep}per_page=${PER_PAGE}&page=${page}`);
+      if (!Array.isArray(batch)) throw new Error(`unexpected ${path} payload: not an array`);
+      yield* batch;
+      if (batch.length < PER_PAGE) return;
+    }
+  }
+
+  return {
+    repo,
+    rest,
+    graphql,
+    list,
+    async closeIssue(issue, comment) {
+      await rest(`/repos/${repo}/issues/${issue}/comments`, {
+        method: "POST",
+        body: JSON.stringify({ body: comment }),
+      });
+      await rest(`/repos/${repo}/issues/${issue}`, {
+        method: "PATCH",
+        body: JSON.stringify({ state: "closed", state_reason: "completed" }),
+      });
+    },
+    async comment(issue, body) {
+      await rest(`/repos/${repo}/issues/${issue}/comments`, {
+        method: "POST",
+        body: JSON.stringify({ body }),
+      });
+    },
+    async react(commentId, content) {
+      await rest(`/repos/${repo}/issues/comments/${commentId}/reactions`, {
+        method: "POST",
+        body: JSON.stringify({ content }),
+      });
+    },
+  };
+}

apps/api/src/index.ts

@@ -0,0 +1,394 @@
+/**
+ * SRS Worker entry: fetch routes + DST-proof cron dispatch.
+ *
+ * Direction of truth: D1 owns SRS state; GitHub issues own the catalog;
+ * the bundled schedule.json owns the calendar. Every handler writes D1
+ * first and runs GitHub/Project side effects afterwards via ctx.waitUntil —
+ * a mirror failure logs a warning and never loses a state write.
+ *
+ * Writes are owner-only (webhook author check, HMAC one-tap links, bearer
+ * admin routes); everything else is read-only public.
+ */
+import { reconcileCatalog } from "./catalog.ts";
+import { gateBadge, heatmapChart, heatmapPng, ladderChart, progressChart } from "./charts.ts";
+import { sendDigest } from "./digest.ts";
+import { createReviewIssue, scoreReview } from "./gate.ts";
+import { type GitHub, github } from "./github.ts";
+import { timingSafeEqual, verifyLink, verifyWebhook } from "./links.ts";
+import { type Mirror, projectMirror } from "./mirror.ts";
+import {
+  CAMPAIGN_START,
+  type LogOutcome,
+  type Stage,
+  daysBetween,
+  dueToday,
+  etDate,
+  etHour,
+  getProblem,
+  logAttempt,
+  weekdayOf,
+} from "./srs.ts";
+import { buildStats } from "./stats.ts";
+
+// ── shared side effects after a state write ──────────────────────
+
+/**
+ * Mirror one logged attempt into GitHub: the Project's Target Date always; on a
+ * first-ever pass also close the problem's sub-issue (comment first — repo
+ * convention). Runs inside ctx.waitUntil; failures are warnings.
+ */
+async function mirrorOutcome(env: Env, outcome: LogOutcome, source: string): Promise<void> {
+  if (outcome.error || outcome.duplicate) return;
+  const gh = github(env.GH_PAT, env.REPO);
+  const mirror = projectMirror(gh, env.REPO.split("/")[0]!);
+  await mirror.setTargetDate(outcome.issue, outcome.next_review);
+  if (outcome.first && outcome.result === "pass") {
+    try {
+      await gh.closeIssue(
+        outcome.issue,
+        `First attempt passed (${outcome.kind}, via ${source}) — entering the review ladder at +3. Logged by the SRS Worker.`,
+      );
+    } catch (err) {
+      console.warn(`close #${outcome.issue}: ${err}`);
+    }
+  }
+}
+
+function outcomeLine(o: LogOutcome): string {
+  if (o.error) return o.error;
+  if (o.duplicate) return `LC ${o.lc}: already logged today — nothing changed`;
+  return (
+    `LC ${o.lc}: ${o.kind} ${o.result} → stage ${o.stage}` +
+    (o.next_review ? `, review ${o.next_review}` : " — retired 🎉")
+  );
+}
+
+// ── one-tap email links ──────────────────────────────────────────
+
+function page(title: string, body: string, status = 200): Response {
+  return new Response(
+    `<!doctype html><meta name="viewport" content="width=device-width,initial-scale=1">
+<body style="font-family:-apple-system,Segoe UI,sans-serif;max-width:420px;margin:15vh auto;padding:0 16px;text-align:center">
+<h2>${title}</h2><p style="color:#57606a">${body}</p></body>`,
+    { status, headers: { "content-type": "text/html; charset=utf-8" } },
+  );
+}
+
+async function handleTap(request: Request, env: Env, ctx: ExecutionContext): Promise<Response> {
+  const url = new URL(request.url);
+  const lc = Number(url.searchParams.get("p"));
+  const result = url.searchParams.get("r");
+  const date = url.searchParams.get("d") ?? "";
+  const sig = url.searchParams.get("sig") ?? "";
+  if (!Number.isFinite(lc) || (result !== "pass" && result !== "fail") || !date) {
+    return page("Bad link", "Missing or malformed parameters.", 400);
+  }
+  if (!(await verifyLink(env.LINK_KEY, lc, result, date, sig))) {
+    return page("Bad signature", "This link was not signed by the SRS.", 403);
+  }
+  const today = etDate(new Date());
+  if (daysBetween(date, today) > 3 || daysBetween(today, date) > 1) {
+    return page("Link expired", "Older than 3 days — log it with a /done comment instead.", 410);
+  }
+
+  const outcome = await logAttempt(env.DB, { lc, date, result, source: "email" });
+  if (outcome.error) return page("Not logged", outcome.error, 422);
+  if (outcome.duplicate) {
+    return page("Already logged ✓", `LC ${lc} was already recorded today. Nothing changed.`);
+  }
+  ctx.waitUntil(mirrorOutcome(env, outcome, "email"));
+  return page(
+    result === "pass" ? "Logged ✅" : "Logged — back to +3",
+    outcomeLine(outcome),
+  );
+}
+
+// ── GitHub webhook ───────────────────────────────────────────────
+
+interface CommentEvent {
+  action: string;
+  issue: { number: number; title: string; labels: { name: string }[] };
+  comment: { id: number; body: string; user: { login: string } };
+}
+
+async function handleWebhook(request: Request, env: Env, ctx: ExecutionContext): Promise<Response> {
+  const raw = await request.arrayBuffer();
+  const ok = await verifyWebhook(env.WEBHOOK_SECRET, raw, request.headers.get("x-hub-signature-256"));
+  if (!ok) return new Response("bad signature", { status: 401 });
+
+  const event = request.headers.get("x-github-event");
+  const payload: unknown = JSON.parse(new TextDecoder().decode(raw));
+  const owner = env.REPO.split("/")[0]!;
+  const today = etDate(new Date());
+
+  if (event === "issues") {
+    if (!payload || typeof payload !== "object" || !("action" in payload) || !("issue" in payload)) {
+      return new Response("ignored");
+    }
+    const p = payload as { action: string; issue: { number: number; labels: { name: string }[] } };
+    if (p.action === "closed" && p.issue.labels.some((l) => l.name === "review")) {
+      const gh = github(env.GH_PAT, env.REPO);
+      const report = await scoreReview(env, gh, p.issue.number, today);
+      return Response.json(report);
+    }
+    return new Response("ignored");
+  }
+
+  if (event !== "issue_comment") return new Response("ignored");
+  const p = payload as CommentEvent;
+  if (p.action !== "created") return new Response("ignored");
+  if (p.comment.user.login !== owner) return new Response("ignored (owner only)");
+
+  const commands: { lc: number; result: "pass" | "fail" }[] = [];
+  const errors: string[] = [];
+  for (const rawLine of p.comment.body.split("\n")) {
+    const line = rawLine.trim();
+    if (!line.startsWith("/done")) continue;
+    const m = line.match(/^\/done\s+(\d+)\s+(pass|fail)\s*$/);
+    if (m) commands.push({ lc: Number(m[1]), result: m[2] as "pass" | "fail" });
+    else errors.push(`cannot parse \`${line}\` — expected \`/done <number> <pass|fail>\``);
+  }
+  if (commands.length === 0 && errors.length === 0) return new Response("no commands");
+
+  // Comments on a review issue log its gate problems as kind='gate'.
+  const gateRow = await env.DB.prepare("SELECT problems FROM gates WHERE issue = ?")
+    .bind(p.issue.number)
+    .first<{ problems: string }>();
+  const gateLcs = new Set<number>(gateRow ? JSON.parse(gateRow.problems) : []);
+
+  const outcomes: LogOutcome[] = [];
+  for (const cmd of commands) {
+    const outcome = await logAttempt(env.DB, {
+      lc: cmd.lc,
+      date: today,
+      result: cmd.result,
+      source: "webhook",
+      gate: gateLcs.has(cmd.lc),
+    });
+    if (outcome.error) errors.push(outcome.error);
+    else outcomes.push(outcome);
+  }
+
+  const gh = github(env.GH_PAT, env.REPO);
+  ctx.waitUntil(
+    (async () => {
+      for (const outcome of outcomes) await mirrorOutcome(env, outcome, "webhook");
+      try {
+        if (errors.length === 0) {
+          await gh.react(p.comment.id, "+1");
+        } else {
+          await gh.comment(
+            p.issue.number,
+            `Could not log everything:\n\n${errors.map((e) => `- ${e}`).join("\n")}` +
+              (outcomes.length
+                ? `\n\nApplied anyway:\n\n${outcomes.map((o) => `- ${outcomeLine(o)}`).join("\n")}`
+                : ""),
+          );
+        }
+      } catch (err) {
+        console.warn(`webhook feedback: ${err}`);
+      }
+    })(),
+  );
+  return Response.json({ applied: outcomes.map(outcomeLine), errors });
+}
+
+// ── admin (bearer LINK_KEY, timing-safe) ─────────────────────────
+
+async function adminAuthorized(request: Request, env: Env): Promise<boolean> {
+  const header = request.headers.get("authorization") ?? "";
+  if (!header.startsWith("Bearer ")) return false;
+  return timingSafeEqual(header.slice(7), env.LINK_KEY);
+}
+
+/**
+ * Solutions landing in `work/` are the third way a problem gets solved, after
+ * the digest tap and a /done comment. `close-solved` posts its whole
+ * implemented set here on every push to main, each entry tagged with the
+ * BUCKET the file sits in — 1 for the first solve, 3 and 7 for the blind
+ * re-solves the picker scaffolds. The bucket names the ladder rung the file
+ * settles, and logAttempt writes only when the problem is standing on that
+ * rung, so this is a total recompute like the reconcilers that call it: a
+ * re-run changes nothing, and a pushed re-solve advances +3 → +7 instead of
+ * being invisible.
+ *
+ * Closing the sub-issue stays with close-solved: it is the side that knows
+ * which files at which commit, so this path mirrors Project fields only.
+ * Charts read D1 live, so a push moves them within the 5-minute cache.
+ */
+const RUNG_OF_BUCKET: Record<number, Stage> = { 1: "new", 3: "+3", 7: "+7" };
+
+async function handleSolved(request: Request, env: Env, date: string): Promise<Response> {
+  const body: unknown = await request.json().catch(() => null);
+  const raw: unknown = body && typeof body === "object" && "solved" in body ? body.solved : null;
+  const shape = 'expected {"solved": [{"lc": <number>, "bucket": 1|3|7}, ...]}';
+  if (!Array.isArray(raw)) return new Response(shape, { status: 400 });
+  const entries: { lc: number; rung: Stage; bucket: number }[] = [];
+  for (const item of raw) {
+    if (!item || typeof item !== "object") return new Response(shape, { status: 400 });
+    const lc: unknown = "lc" in item ? item.lc : null;
+    const bucket: unknown = "bucket" in item ? item.bucket : null;
+    if (typeof lc !== "number" || !Number.isFinite(lc)) {
+      return new Response(`not an LC number: ${JSON.stringify(lc)}`, { status: 400 });
+    }
+    const rung = typeof bucket === "number" ? RUNG_OF_BUCKET[bucket] : undefined;
+    if (rung === undefined) return new Response(`not a work/ bucket: ${JSON.stringify(bucket)}`, { status: 400 });
+    entries.push({ lc, rung, bucket });
+  }
+  const dry = new URL(request.url).searchParams.get("dry") === "1";
+
+  const logged: string[] = [];
+  const skipped: string[] = [];
+  let mirror: Mirror | undefined;
+  for (const { lc, rung, bucket } of entries) {
+    if (dry) {
+      const p = await getProblem(env.DB, lc);
+      if (!p) skipped.push(`LC ${lc} is not in the curriculum`);
+      else if (p.stage !== rung) skipped.push(`LC ${lc}: work/${bucket} settles ${rung}, stage is ${p.stage}`);
+      else if (rung === "new") logged.push(`LC ${lc}: would enter the ladder at +3`);
+      else logged.push(`LC ${lc}: would pass ${rung} and move on`);
+      continue;
+    }
+    const outcome = await logAttempt(env.DB, { lc, date, result: "pass", source: "commit", rung });
+    if (outcome.error) skipped.push(outcome.error);
+    else if (outcome.duplicate) {
+      skipped.push(`LC ${lc}: work/${bucket} settles ${rung}, stage is ${outcome.stage}`);
+    } else {
+      logged.push(outcomeLine(outcome));
+      // One mirror for the batch: field IDs resolve once, not per problem.
+      mirror ??= projectMirror(github(env.GH_PAT, env.REPO), env.REPO.split("/")[0]!);
+      await mirror.setTargetDate(outcome.issue, outcome.next_review);
+    }
+  }
+  return Response.json({ date, logged, skipped, warnings: mirror?.warnings ?? [] });
+}
+
+async function handleAdmin(request: Request, env: Env, path: string): Promise<Response> {
+  if (!(await adminAuthorized(request, env))) return new Response("unauthorized", { status: 401 });
+  const url = new URL(request.url);
+  const date = url.searchParams.get("date") ?? etDate(new Date());
+  const gh = () => github(env.GH_PAT, env.REPO);
+
+  if (path === "/admin/reconcile") {
+    return Response.json(await reconcileCatalog(env.DB, gh()));
+  }
+  if (path === "/admin/solved") return await handleSolved(request, env, date);
+  if (path === "/admin/digest") {
+    const report = await sendDigest(env, date, {
+      dry: url.searchParams.get("dry") === "1",
+      force: url.searchParams.get("force") === "1",
+    });
+    if (url.searchParams.get("dry") === "1") {
+      return new Response(report.digest.html, { headers: { "content-type": "text/html" } });
+    }
+    return Response.json({ sent: report.sent, reason: report.reason, subject: report.digest.subject });
+  }
+  if (path === "/admin/review") {
+    return Response.json(await createReviewIssue(env, gh(), date));
+  }
+  // What is due on `date`, straight out of the ladder — the ONE answer to
+  // "which problems are open today", so the daily review issue and the digest
+  // cannot disagree. Read-only; the caller decides what to render.
+  if (path === "/admin/due") {
+    return Response.json(await dueToday(env.DB, date));
+  }
+  return new Response("not found", { status: 404 });
+}
+
+// ── router ───────────────────────────────────────────────────────
+
+const SVG_HEADERS = {
+  "content-type": "image/svg+xml",
+  // GitHub Camo honors this; 5 minutes is the freshness floor for README charts.
+  "cache-control": "public, max-age=300",
+};
+
+// The digest embeds the raster heatmap; email clients refuse SVG entirely.
+const PNG_HEADERS = {
+  "content-type": "image/png",
+  "cache-control": "public, max-age=300",
+};
+
+export default {
+  async fetch(request: Request, env: Env, ctx: ExecutionContext): Promise<Response> {
+    const url = new URL(request.url);
+    const path = url.pathname;
+
+    try {
+      if (path === "/log" && request.method === "GET") return await handleTap(request, env, ctx);
+      if (path === "/webhook/github" && request.method === "POST") {
+        return await handleWebhook(request, env, ctx);
+      }
+      if (path.startsWith("/admin/") && request.method === "POST") {
+        return await handleAdmin(request, env, path);
+      }
+
+      if (path === "/chart/progress.svg") {
+        return new Response(await progressChart(env.DB), { headers: SVG_HEADERS });
+      }
+      if (path === "/chart/ladder.svg") {
+        return new Response(await ladderChart(env.DB), { headers: SVG_HEADERS });
+      }
+      if (path === "/chart/heatmap.svg") {
+        return new Response(await heatmapChart(env.DB, CAMPAIGN_START, 56), { headers: SVG_HEADERS });
+      }
+      if (path === "/chart/heatmap.png") {
+        return new Response(await heatmapPng(env.DB, CAMPAIGN_START, 56), { headers: PNG_HEADERS });
+      }
+      if (path === "/badge/gate.svg") {
+        return new Response(await gateBadge(env.DB), { headers: SVG_HEADERS });
+      }
+
+      if (path === "/api/stats") {
+        // The docs serve from both the Pages origin and the custom domain;
+        // reflect the requesting origin only when it is on the allowlist.
+        const allowed = env.DOCS_ORIGIN.split(",").map((o) => o.trim());
+        const origin = request.headers.get("origin");
+        const cors = {
+          "access-control-allow-origin":
+            origin && allowed.includes(origin) ? origin : allowed[0]!,
+          "access-control-allow-methods": "GET",
+          vary: "Origin",
+        };
+        if (request.method === "OPTIONS") return new Response(null, { headers: cors });
+        const stats = await buildStats(env.DB, etDate(new Date()));
+        return Response.json(stats, { headers: cors });
+      }
+
+      return new Response("srs-api", { status: path === "/" ? 200 : 404 });
+    } catch (err) {
+      console.error(`unhandled ${request.method} ${path}: ${err}`);
+      return new Response("internal error", { status: 500 });
+    }
+  },
+
+  /**
+   * DST-proof cron dispatch: crons fire at both possible UTC hours; the
+   * computed ET hour decides. 8 AM daily → catalog reconcile + digest;
+   * midnight Saturday → review issue (so the 8 AM digest can link to it).
+   */
+  async scheduled(controller: ScheduledController, env: Env, ctx: ExecutionContext): Promise<void> {
+    const now = new Date(controller.scheduledTime);
+    const hour = etHour(now);
+    const date = etDate(now);
+
+    if (hour === 8) {
+      try {
+        const report = await reconcileCatalog(env.DB, github(env.GH_PAT, env.REPO));
+        console.log(`catalog: ${report.topics} topics, ${report.problems} problems`);
+      } catch (err) {
+        console.warn(`catalog reconcile failed (digest still goes out): ${err}`);
+      }
+      const report = await sendDigest(env, date);
+      console.log(`digest ${date}: ${report.reason}`);
+      return;
+    }
+    if (hour === 0 && weekdayOf(date) === 6) {
+      const report = await createReviewIssue(env, github(env.GH_PAT, env.REPO), date);
+      console.log(`review issue ${date}: ${report.reason}`);
+      return;
+    }
+    console.log(`cron at ET hour ${hour} on ${date}: no-op (DST guard)`);
+    void ctx;
+  },
+};

apps/api/src/links.ts

@@ -0,0 +1,79 @@
+/**
+ * One-tap email links: `GET /log?p=704&r=pass&d=2026-08-31&sig=<hmac>`.
+ * The sig is HMAC-SHA256 over `p|r|d` with LINK_KEY (WebCrypto, hex).
+ * Verification is constant-time; links older than 3 days are rejected.
+ */
+
+async function hmacKey(secret: string): Promise<CryptoKey> {
+  return crypto.subtle.importKey(
+    "raw",
+    new TextEncoder().encode(secret),
+    { name: "HMAC", hash: "SHA-256" },
+    false,
+    ["sign", "verify"],
+  );
+}
+
+export async function signLink(
+  key: string,
+  p: number,
+  r: string,
+  d: string,
+): Promise<string> {
+  const mac = await crypto.subtle.sign(
+    "HMAC",
+    await hmacKey(key),
+    new TextEncoder().encode(`${p}|${r}|${d}`),
+  );
+  return [...new Uint8Array(mac)].map((b) => b.toString(16).padStart(2, "0")).join("");
+}
+
+export async function verifyLink(
+  key: string,
+  p: number,
+  r: string,
+  d: string,
+  sig: string,
+): Promise<boolean> {
+  if (!/^[0-9a-f]{64}$/.test(sig)) return false;
+  const bytes = new Uint8Array(32);
+  for (let i = 0; i < 32; i++) bytes[i] = Number.parseInt(sig.slice(i * 2, i * 2 + 2), 16);
+  // crypto.subtle.verify is constant-time; never compare hex strings directly.
+  return crypto.subtle.verify(
+    "HMAC",
+    await hmacKey(key),
+    bytes,
+    new TextEncoder().encode(`${p}|${r}|${d}`),
+  );
+}
+
+/** Constant-time equality for webhook signatures and admin bearer keys. */
+export async function timingSafeEqual(a: string, b: string): Promise<boolean> {
+  // HMAC both sides with a random key: unequal-length inputs and content
+  // differences are equally invisible to timing.
+  const key = await crypto.subtle.generateKey({ name: "HMAC", hash: "SHA-256" }, false, [
+    "sign",
+  ]);
+  const enc = new TextEncoder();
+  const [ma, mb] = await Promise.all([
+    crypto.subtle.sign("HMAC", key, enc.encode(a)),
+    crypto.subtle.sign("HMAC", key, enc.encode(b)),
+  ]);
+  const va = new Uint8Array(ma);
+  const vb = new Uint8Array(mb);
+  let diff = 0;
+  for (let i = 0; i < va.length; i++) diff |= va[i]! ^ vb[i]!;
+  return diff === 0;
+}
+
+/** GitHub webhook `X-Hub-Signature-256: sha256=<hex>` verification. */
+export async function verifyWebhook(
+  secret: string,
+  body: ArrayBuffer,
+  header: string | null,
+): Promise<boolean> {
+  if (!header?.startsWith("sha256=")) return false;
+  const mac = await crypto.subtle.sign("HMAC", await hmacKey(secret), body);
+  const expected = [...new Uint8Array(mac)].map((b) => b.toString(16).padStart(2, "0")).join("");
+  return timingSafeEqual(header.slice(7), expected);
+}

apps/api/src/mirror.ts

@@ -0,0 +1,143 @@
+/**
+ * Best-effort mirror into the "Interview Prep" user Project — ported from
+ * scripts/srs-project.ts. D1 is the truth; a GraphQL failure here becomes a
+ * console warning and never blocks a state write. Only problem issues are
+ * ever passed in, so topic rows' Target Date is never written.
+ *
+ * Target Date is the ONLY mirrored column. SRS stage and first-attempt result
+ * live exclusively in D1 — they are read from there (charts, digest, /api/stats)
+ * and duplicating them into Project single-selects bought nothing but drift and
+ * two extra writes per log. Do not re-add them here.
+ *
+ * Field IDs are resolved by NAME per invocation (a Worker isolate is
+ * short-lived; the extra query per mirror burst is irrelevant at this
+ * volume and survives field re-creation).
+ */
+import type { GitHub } from "./github.ts";
+
+const PROJECT_TITLE = "Interview Prep";
+
+interface ProjectInfo {
+  id: string;
+  targetDate: string;
+}
+
+export interface Mirror {
+  setTargetDate(issue: number, date: string | null): Promise<void>;
+  warnings: string[];
+}
+
+export function projectMirror(gh: GitHub, owner: string): Mirror {
+  const warnings: string[] = [];
+  let info: ProjectInfo | undefined;
+  const itemIds = new Map<number, string | undefined>();
+
+  async function resolve(): Promise<ProjectInfo> {
+    if (info) return info;
+    const data = (await gh.graphql(
+      `query($owner: String!, $title: String!) {
+        user(login: $owner) {
+          projectsV2(first: 10, query: $title) {
+            nodes {
+              id title
+              fields(first: 30) {
+                nodes {
+                  ... on ProjectV2FieldCommon { id name dataType }
+                }
+              }
+            }
+          }
+        }
+      }`,
+      { owner, title: PROJECT_TITLE },
+    )) as {
+      user: {
+        projectsV2: {
+          nodes: {
+            id: string;
+            title: string;
+            fields: {
+              nodes: { id: string; name: string }[];
+            };
+          }[];
+        };
+      };
+    };
+    const node = data.user.projectsV2.nodes.find((n) => n.title === PROJECT_TITLE);
+    if (!node) throw new Error(`project "${PROJECT_TITLE}" not found for @${owner}`);
+    const date = node.fields.nodes.find((n) => n.name === "Target Date");
+    if (!date) throw new Error(`date field "Target Date" missing`);
+    info = { id: node.id, targetDate: date.id };
+    return info;
+  }
+
+  async function itemId(issue: number): Promise<string> {
+    if (itemIds.has(issue)) {
+      const cached = itemIds.get(issue);
+      if (!cached) throw new Error(`issue #${issue} is not in the project`);
+      return cached;
+    }
+    const project = await resolve();
+    const [repoOwner, repoName] = gh.repo.split("/");
+    const data = (await gh.graphql(
+      `query($owner: String!, $name: String!, $issue: Int!) {
+        repository(owner: $owner, name: $name) {
+          issue(number: $issue) {
+            projectItems(first: 10, includeArchived: true) {
+              nodes { id project { id } }
+            }
+          }
+        }
+      }`,
+      { owner: repoOwner, name: repoName, issue },
+    )) as {
+      repository: { issue: { projectItems: { nodes: { id: string; project: { id: string } }[] } } };
+    };
+    const item = data.repository.issue.projectItems.nodes.find((n) => n.project.id === project.id);
+    itemIds.set(issue, item?.id);
+    if (!item) throw new Error(`issue #${issue} is not in the project`);
+    return item.id;
+  }
+
+  async function setField(issue: number, fieldId: string, value: object): Promise<void> {
+    const project = await resolve();
+    await gh.graphql(
+      `mutation($project: ID!, $item: ID!, $field: ID!, $value: ProjectV2FieldValue!) {
+        updateProjectV2ItemFieldValue(
+          input: { projectId: $project, itemId: $item, fieldId: $field, value: $value }
+        ) { projectV2Item { id } }
+      }`,
+      { project: project.id, item: await itemId(issue), field: fieldId, value },
+    );
+  }
+
+  async function attempt(what: string, op: () => Promise<void>): Promise<void> {
+    try {
+      await op();
+    } catch (err) {
+      const message = `mirror ${what}: ${err instanceof Error ? err.message : String(err)}`;
+      warnings.push(message);
+      console.warn(message); // observability picks this up; never rethrow
+    }
+  }
+
+  return {
+    warnings,
+    setTargetDate: (issue, date) =>
+      attempt(`Target Date #${issue}`, async () => {
+        const project = await resolve();
+        if (date === null) {
+          await gh.graphql(
+            `mutation($project: ID!, $item: ID!, $field: ID!) {
+              clearProjectV2ItemFieldValue(
+                input: { projectId: $project, itemId: $item, fieldId: $field }
+              ) { projectV2Item { id } }
+            }`,
+            { project: project.id, item: await itemId(issue), field: project.targetDate },
+          );
+        } else {
+          await setField(issue, project.targetDate, { date });
+        }
+      }),
+  };
+}

apps/api/src/png.ts

@@ -0,0 +1,254 @@
+/**
+ * Hand-rolled PNG encoder for the SRS Worker.
+ *
+ * Every major email client — Gmail, Outlook, Apple Mail — refuses to render
+ * SVG, inline or remote, so the daily digest cannot reuse /chart/heatmap.svg;
+ * it needs raster bytes. The Worker ships with no runtime dependencies by
+ * design (SVG, HMAC and GraphQL are all hand-rolled here), so this file
+ * encodes PNG itself instead of pulling in a codec.
+ *
+ * The format reduces to four things the runtime already provides:
+ *   - an 8-byte signature plus chunks (IHDR, IDAT, IEND), each one
+ *     length + type + payload + CRC32, every integer big-endian;
+ *   - one filter byte per scanline, always 0 here: these images are flat
+ *     rectangles, so a predictor would buy nothing;
+ *   - zlib-wrapped deflate for IDAT, which is exactly what
+ *     `new CompressionStream("deflate")` emits ("deflate-raw" would not);
+ *   - CRC32, a 256-entry table built lazily on the first chunk.
+ *
+ * Colour type 2 (truecolour, 8-bit, no alpha): these charts are opaque cards,
+ * so an alpha channel would cost a third more bytes for nothing. That does mean
+ * transparency is unavailable — a caller wanting rounded corners paints the
+ * colour that sits behind the image first, then the rounded card on top.
+ */
+
+// ── crc32 ────────────────────────────────────────────────────────
+
+// One table for every chunk of every image; built on first use.
+let crcTable: Uint32Array | undefined;
+
+function crc32(bytes: Uint8Array): number {
+  let table = crcTable;
+  if (!table) {
+    table = new Uint32Array(256);
+    for (let n = 0; n < 256; n++) {
+      let c = n;
+      for (let k = 0; k < 8; k++) c = (c & 1) !== 0 ? 0xedb88320 ^ (c >>> 1) : c >>> 1;
+      table[n] = c;
+    }
+    crcTable = table;
+  }
+  let crc = 0xffffffff;
+  for (const byte of bytes) crc = table[(crc ^ byte) & 0xff]! ^ (crc >>> 8);
+  return (crc ^ 0xffffffff) >>> 0;
+}
+
+// ── chunks ───────────────────────────────────────────────────────
+
+const SIGNATURE = new Uint8Array([0x89, 0x50, 0x4e, 0x47, 0x0d, 0x0a, 0x1a, 0x0a]);
+
+/** length, 4-char type, payload, then CRC32 over the type and payload. */
+function chunk(type: string, data: Uint8Array): Uint8Array {
+  const out = new Uint8Array(data.length + 12);
+  const view = new DataView(out.buffer);
+  view.setUint32(0, data.length);
+  for (let i = 0; i < 4; i++) out[4 + i] = type.charCodeAt(i);
+  out.set(data, 8);
+  view.setUint32(data.length + 8, crc32(out.subarray(4, data.length + 8)));
+  return out;
+}
+
+// ── 5x7 bitmap font ──────────────────────────────────────────────
+
+const GLYPH_W = 5;
+const GLYPH_H = 7;
+
+/**
+ * Seven row bitmasks per glyph, five low bits each, MSB (0b10000) leftmost.
+ * Uppercase only: chart labels are short and mechanical, and a lowercase set
+ * would double the table for no gain. Unlisted characters render blank.
+ */
+const GLYPHS: Record<string, number[]> = {
+  A: [0b01110, 0b10001, 0b10001, 0b11111, 0b10001, 0b10001, 0b10001],
+  B: [0b11110, 0b10001, 0b10001, 0b11110, 0b10001, 0b10001, 0b11110],
+  C: [0b01110, 0b10001, 0b10000, 0b10000, 0b10000, 0b10001, 0b01110],
+  D: [0b11110, 0b10001, 0b10001, 0b10001, 0b10001, 0b10001, 0b11110],
+  E: [0b11111, 0b10000, 0b10000, 0b11110, 0b10000, 0b10000, 0b11111],
+  F: [0b11111, 0b10000, 0b10000, 0b11110, 0b10000, 0b10000, 0b10000],
+  G: [0b01110, 0b10001, 0b10000, 0b10111, 0b10001, 0b10001, 0b01110],
+  H: [0b10001, 0b10001, 0b10001, 0b11111, 0b10001, 0b10001, 0b10001],
+  I: [0b11111, 0b00100, 0b00100, 0b00100, 0b00100, 0b00100, 0b11111],
+  J: [0b00111, 0b00010, 0b00010, 0b00010, 0b00010, 0b10010, 0b01100],
+  K: [0b10001, 0b10010, 0b10100, 0b11000, 0b10100, 0b10010, 0b10001],
+  L: [0b10000, 0b10000, 0b10000, 0b10000, 0b10000, 0b10000, 0b11111],
+  M: [0b10001, 0b11011, 0b10101, 0b10101, 0b10001, 0b10001, 0b10001],
+  N: [0b10001, 0b10001, 0b11001, 0b10101, 0b10011, 0b10001, 0b10001],
+  O: [0b01110, 0b10001, 0b10001, 0b10001, 0b10001, 0b10001, 0b01110],
+  P: [0b11110, 0b10001, 0b10001, 0b11110, 0b10000, 0b10000, 0b10000],
+  Q: [0b01110, 0b10001, 0b10001, 0b10001, 0b10101, 0b10011, 0b01101],
+  R: [0b11110, 0b10001, 0b10001, 0b11110, 0b10100, 0b10010, 0b10001],
+  S: [0b01111, 0b10000, 0b10000, 0b01110, 0b00001, 0b00001, 0b11110],
+  T: [0b11111, 0b00100, 0b00100, 0b00100, 0b00100, 0b00100, 0b00100],
+  U: [0b10001, 0b10001, 0b10001, 0b10001, 0b10001, 0b10001, 0b01110],
+  V: [0b10001, 0b10001, 0b10001, 0b10001, 0b10001, 0b01010, 0b00100],
+  W: [0b10001, 0b10001, 0b10001, 0b10101, 0b10101, 0b11011, 0b10001],
+  X: [0b10001, 0b10001, 0b01010, 0b00100, 0b01010, 0b10001, 0b10001],
+  Y: [0b10001, 0b10001, 0b01010, 0b00100, 0b00100, 0b00100, 0b00100],
+  Z: [0b11111, 0b00001, 0b00010, 0b00100, 0b01000, 0b10000, 0b11111],
+  "0": [0b01110, 0b10001, 0b10011, 0b10101, 0b11001, 0b10001, 0b01110],
+  "1": [0b00100, 0b01100, 0b00100, 0b00100, 0b00100, 0b00100, 0b01110],
+  "2": [0b01110, 0b10001, 0b00001, 0b00010, 0b00100, 0b01000, 0b11111],
+  "3": [0b11111, 0b00010, 0b00100, 0b00010, 0b00001, 0b10001, 0b01110],
+  "4": [0b00010, 0b00110, 0b01010, 0b10010, 0b11111, 0b00010, 0b00010],
+  "5": [0b11111, 0b10000, 0b11110, 0b00001, 0b00001, 0b10001, 0b01110],
+  "6": [0b00110, 0b01000, 0b10000, 0b11110, 0b10001, 0b10001, 0b01110],
+  "7": [0b11111, 0b00001, 0b00010, 0b00100, 0b01000, 0b01000, 0b01000],
+  "8": [0b01110, 0b10001, 0b10001, 0b01110, 0b10001, 0b10001, 0b01110],
+  "9": [0b01110, 0b10001, 0b10001, 0b01111, 0b00001, 0b00010, 0b01100],
+  " ": [0b00000, 0b00000, 0b00000, 0b00000, 0b00000, 0b00000, 0b00000],
+  "-": [0b00000, 0b00000, 0b00000, 0b11111, 0b00000, 0b00000, 0b00000],
+  "/": [0b00001, 0b00001, 0b00010, 0b00100, 0b01000, 0b10000, 0b10000],
+  ":": [0b00000, 0b00100, 0b00100, 0b00000, 0b00100, 0b00100, 0b00000],
+};
+
+/**
+ * Device pixels `content` occupies at `scale`, inter-glyph gaps included.
+ * Exported because callers need it to size a canvas before constructing one.
+ */
+export function textWidth(content: string, scale = 1): number {
+  return content.length === 0 ? 0 : (content.length * (GLYPH_W + 1) - 1) * scale;
+}
+
+// ── canvas ───────────────────────────────────────────────────────
+
+/** "#rrggbb" to its three channel bytes; called once per draw, not per pixel. */
+function rgb(color: string): [number, number, number] {
+  const packed = Number.parseInt(color.slice(1), 16);
+  return [(packed >> 16) & 0xff, (packed >> 8) & 0xff, packed & 0xff];
+}
+
+/** Fixed-size RGB canvas: filled rectangles plus 5x7 bitmap text. */
+export class Canvas {
+  readonly width: number;
+  readonly height: number;
+  /** Row-major RGB triples, unpadded; encode() interleaves the filter bytes. */
+  private readonly pixels: Uint8Array;
+
+  constructor(width: number, height: number, background: string) {
+    this.width = width;
+    this.height = height;
+    this.pixels = new Uint8Array(width * height * 3);
+    const [r, g, b] = rgb(background);
+    for (let i = 0; i < this.pixels.length; i += 3) {
+      this.pixels[i] = r;
+      this.pixels[i + 1] = g;
+      this.pixels[i + 2] = b;
+    }
+  }
+
+  /**
+   * Filled rectangle, clipped to the canvas. `radius` rounds all four corners
+   * by dropping the pixels whose centre falls outside the corner circle —
+   * enough for the small cell roundings and the card these charts draw.
+   */
+  rect(x: number, y: number, w: number, h: number, color: string, radius = 0): void {
+    const [r, g, b] = rgb(color);
+    // Device pixels only: a fractional origin would index the buffer between
+    // bytes and silently drop the whole rectangle.
+    const left = Math.round(x);
+    const top = Math.round(y);
+    const rw = Math.round(w);
+    const rh = Math.round(h);
+    const rad = Math.min(radius, rw / 2, rh / 2);
+
+    const y1 = Math.min(this.height, top + rh);
+    const x1 = Math.min(this.width, left + rw);
+    for (let py = Math.max(0, top); py < y1; py++) {
+      // How far this row lies past the nearer corner centre; 0 in between.
+      const cy = py + 0.5;
+      const dy = cy < top + rad ? top + rad - cy : cy > top + rh - rad ? cy - (top + rh - rad) : 0;
+      const rowBase = py * this.width * 3;
+      for (let px = Math.max(0, left); px < x1; px++) {
+        const cx = px + 0.5;
+        const dx = cx < left + rad ? left + rad - cx : cx > left + rw - rad ? cx - (left + rw - rad) : 0;
+        if (dx * dx + dy * dy > rad * rad) continue;
+        const at = rowBase + px * 3;
+        this.pixels[at] = r;
+        this.pixels[at + 1] = g;
+        this.pixels[at + 2] = b;
+      }
+    }
+  }
+
+  /**
+   * Bitmap text from a top-left origin (not a baseline), `scale` device pixels
+   * per glyph pixel, 1 glyph pixel of advance between characters. Input is
+   * uppercased; unknown characters advance without drawing.
+   */
+  text(x: number, y: number, content: string, color: string, scale = 1): void {
+    const [r, g, b] = rgb(color);
+    const left = Math.round(x);
+    const top = Math.round(y);
+    const upper = content.toUpperCase();
+
+    for (let i = 0; i < upper.length; i++) {
+      const glyph = GLYPHS[upper[i]!];
+      if (!glyph) continue;
+      const originX = left + i * (GLYPH_W + 1) * scale;
+      for (let gy = 0; gy < GLYPH_H; gy++) {
+        const bits = glyph[gy]!;
+        if (bits === 0) continue;
+        for (let gx = 0; gx < GLYPH_W; gx++) {
+          if ((bits & (1 << (GLYPH_W - 1 - gx))) === 0) continue;
+          // Every set glyph pixel is a scale x scale block of device pixels.
+          const blockX = originX + gx * scale;
+          const blockY = top + gy * scale;
+          const y1 = Math.min(this.height, blockY + scale);
+          const x1 = Math.min(this.width, blockX + scale);
+          for (let py = Math.max(0, blockY); py < y1; py++) {
+            const rowBase = py * this.width * 3;
+            for (let px = Math.max(0, blockX); px < x1; px++) {
+              const at = rowBase + px * 3;
+              this.pixels[at] = r;
+              this.pixels[at + 1] = g;
+              this.pixels[at + 2] = b;
+            }
+          }
+        }
+      }
+    }
+  }
+
+  /** PNG bytes: 8-bit truecolour, one IDAT, zlib via CompressionStream. */
+  async encode(): Promise<Uint8Array> {
+    const stride = this.width * 3;
+    const raw = new Uint8Array((stride + 1) * this.height);
+    for (let y = 0; y < this.height; y++) {
+      // Leading 0 of each scanline slot is the "no filter" byte.
+      raw.set(this.pixels.subarray(y * stride, (y + 1) * stride), y * (stride + 1) + 1);
+    }
+    const deflated = new Response(raw).body!.pipeThrough(new CompressionStream("deflate"));
+    const idat = new Uint8Array(await new Response(deflated).arrayBuffer());
+
+    const ihdr = new Uint8Array(13);
+    const header = new DataView(ihdr.buffer);
+    header.setUint32(0, this.width);
+    header.setUint32(4, this.height);
+    ihdr[8] = 8; // bit depth
+    ihdr[9] = 2; // colour type 2: truecolour RGB
+    // Bytes 10-12 stay 0: deflate, adaptive filtering, no interlace.
+
+    const chunks = [chunk("IHDR", ihdr), chunk("IDAT", idat), chunk("IEND", new Uint8Array(0))];
+    let total = SIGNATURE.length;
+    for (const c of chunks) total += c.length;
+
+    const png = new Uint8Array(total);
+    png.set(SIGNATURE);
+    let at = SIGNATURE.length;
+    for (const c of chunks) {
+      png.set(c, at);
+      at += c.length;
+    }
+    return png;
+  }
+}

apps/api/src/srs.test.ts

@@ -0,0 +1,168 @@
+/**
+ * Acceptance proof for the two rules that cannot be eyeballed in production:
+ * the double crons fire exactly once at 8 AM ET (digest) and midnight Saturday
+ * ET (review issue) on BOTH UTC offsets, and NOTHING is ever scheduled on the
+ * Sunday rest day — which is the whole reason the ladder's windows are 3 and 7.
+ */
+import { describe, expect, test } from "bun:test";
+
+import {
+  INTERVAL,
+  STAGES,
+  WINDOWS,
+  addDays,
+  campaignDay,
+  campaignWeek,
+  etDate,
+  etHour,
+  isRestDay,
+  isoWeek,
+  rng,
+  sample,
+  TEMPERATURES,
+  temperatureOf,
+  weekdayOf,
+  workingDay,
+} from "./srs.ts";
+
+/** Every ET date of the campaign, plus the deferred-Hard tail. */
+const CALENDAR = Array.from({ length: 70 }, (_, i) => addDays("2026-08-17", i));
+
+describe("DST-proof cron guard", () => {
+  test("digest fires only at ET hour 8 — EDT (UTC-4)", () => {
+    // Cron pair 12:00 / 13:00 UTC during EDT (2026-08-29 is EDT).
+    expect(etHour(new Date("2026-08-29T12:00:00Z"))).toBe(8); // fires
+    expect(etHour(new Date("2026-08-29T13:00:00Z"))).toBe(9); // no-op
+  });
+  test("digest fires only at ET hour 8 — EST (UTC-5)", () => {
+    // DST ends 2026-11-01; 2026-11-05 is EST.
+    expect(etHour(new Date("2026-11-05T12:00:00Z"))).toBe(7); // no-op
+    expect(etHour(new Date("2026-11-05T13:00:00Z"))).toBe(8); // fires
+  });
+  test("saturday review fires only at ET midnight — both offsets", () => {
+    // EDT Saturday: 4 UTC = 0 ET fires, 5 UTC = 1 ET no-op.
+    expect(etHour(new Date("2026-08-29T04:00:00Z"))).toBe(0);
+    expect(weekdayOf(etDate(new Date("2026-08-29T04:00:00Z")))).toBe(6);
+    expect(etHour(new Date("2026-08-29T05:00:00Z"))).toBe(1);
+    // EST Saturday (2026-11-07): 5 UTC = 0 ET fires, 4 UTC = 11 PM FRIDAY ET.
+    expect(etHour(new Date("2026-11-07T05:00:00Z"))).toBe(0);
+    expect(weekdayOf(etDate(new Date("2026-11-07T05:00:00Z")))).toBe(6);
+    expect(etHour(new Date("2026-11-07T04:00:00Z"))).toBe(23);
+    expect(etDate(new Date("2026-11-07T04:00:00Z"))).toBe("2026-11-06"); // still Friday ET
+  });
+});
+
+describe("date math (noon-UTC anchored)", () => {
+  test("addDays crosses the DST-end boundary without drift", () => {
+    expect(addDays("2026-10-31", 2)).toBe("2026-11-02");
+    expect(addDays("2026-08-24", 2)).toBe("2026-08-26");
+  });
+  test("campaign math", () => {
+    expect(campaignDay("2026-08-17")).toBe(1);
+    expect(campaignDay("2026-08-24")).toBe(8);
+    expect(campaignWeek("2026-08-24")).toBe(2);
+    expect(campaignWeek("2026-08-30")).toBe(2);
+    expect(campaignWeek("2026-08-31")).toBe(3);
+  });
+  test("iso week is stable across a week", () => {
+    expect(isoWeek("2026-08-24")).toBe(isoWeek("2026-08-29"));
+    expect(isoWeek("2026-08-30")).toBe(isoWeek("2026-08-24")); // Sun ends the ISO week
+    expect(isoWeek("2026-08-31")).toBe(isoWeek("2026-08-24") + 1);
+  });
+});
+
+describe("the rest day", () => {
+  test("Sunday is the only rest day", () => {
+    expect(isRestDay("2026-08-30")).toBe(true); // Sunday
+    expect(isRestDay("2026-08-29")).toBe(false); // Saturday: gate day, still work
+    expect(isRestDay("2026-08-31")).toBe(false);
+  });
+
+  test("workingDay never returns a Sunday, from any day and any offset", () => {
+    for (const date of CALENDAR) {
+      for (let offset = 0; offset < 10; offset++) {
+        expect(isRestDay(workingDay(date, offset))).toBe(false);
+      }
+    }
+  });
+
+  test("workingDay(date) leaves a working day alone and slides Sunday to Monday", () => {
+    expect(workingDay("2026-09-04")).toBe("2026-09-04"); // Friday
+    expect(workingDay("2026-09-05")).toBe("2026-09-05"); // Saturday
+    expect(workingDay("2026-09-06")).toBe("2026-09-07"); // Sunday → Monday
+  });
+
+  test("offsets count working days, so a spill skips the rest day", () => {
+    // Levelling from Saturday: tomorrow is Sunday, so the spill starts Monday.
+    expect(workingDay("2026-09-05", 1)).toBe("2026-09-07");
+    expect(workingDay("2026-09-05", 2)).toBe("2026-09-08");
+    // Two deferred Hards per working day from a Monday.
+    expect(workingDay("2026-09-28", 5)).toBe("2026-10-03"); // Saturday
+    expect(workingDay("2026-09-28", 6)).toBe("2026-10-05"); // Sunday skipped
+  });
+
+  test("every ladder interval lands on a working day, from every learning day", () => {
+    for (const date of CALENDAR) {
+      if (isRestDay(date)) continue;
+      for (const days of WINDOWS) {
+        const review = workingDay(addDays(date, days));
+        expect(isRestDay(review)).toBe(false);
+        // Never earlier than the interval: a review may slip, never shorten.
+        expect(review >= addDays(date, days)).toBe(true);
+      }
+    }
+  });
+
+  test("only Thursday's +3 needs the slide", () => {
+    const slid = CALENDAR.filter(
+      (date) => !isRestDay(date) && workingDay(addDays(date, 3)) !== addDays(date, 3),
+    );
+    expect(new Set(slid.map(weekdayOf))).toEqual(new Set([4])); // Thursday only
+    for (const date of CALENDAR) {
+      if (isRestDay(date)) continue;
+      expect(workingDay(addDays(date, 7))).toBe(addDays(date, 7)); // +7 keeps the weekday
+    }
+  });
+});
+
+describe("the ladder", () => {
+  test("two rungs, named after the windows", () => {
+    expect(STAGES).toEqual(["new", "+3", "+7", "retired"]);
+    expect(WINDOWS.map((w) => INTERVAL[`+${w}`])).toEqual([...WINDOWS]);
+    // Every non-terminal stage has an interval, and only those.
+    expect(Object.keys(INTERVAL).sort()).toEqual(STAGES.slice(1, -1).sort());
+  });
+});
+
+describe("concept temperature", () => {
+  test("the bands are the ladder's own windows", () => {
+    // Boundaries are inclusive, and the bands tile the whole age axis with no
+    // gap: whatever the windows become, every age still lands in exactly one.
+    expect(TEMPERATURES).toEqual(["hot", "fresh", "fading", "cold"]);
+    expect(temperatureOf(0)).toBe("hot");
+    expect(temperatureOf(WINDOWS[0])).toBe("hot");
+    expect(temperatureOf(WINDOWS[0] + 1)).toBe("fresh");
+    expect(temperatureOf(WINDOWS[1])).toBe("fresh");
+    expect(temperatureOf(WINDOWS[1] + 1)).toBe("fading");
+    expect(temperatureOf(WINDOWS[1] * 2)).toBe("fading");
+    expect(temperatureOf(WINDOWS[1] * 2 + 1)).toBe("cold");
+  });
+
+  test("never attempted is cold, not unknown", () => {
+    expect(temperatureOf(null)).toBe("cold");
+  });
+
+  test("every age of the campaign has exactly one band", () => {
+    for (let age = 0; age <= CALENDAR.length; age++) {
+      expect(TEMPERATURES).toContain(temperatureOf(age));
+    }
+  });
+});
+
+describe("deterministic sampling", () => {
+  test("same seed, same picks", () => {
+    const pool = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10];
+    expect(sample(pool, 3, rng("drill-2026-08-31"))).toEqual(sample(pool, 3, rng("drill-2026-08-31")));
+    expect(sample(pool, 3, rng("drill-2026-09-01"))).not.toEqual(sample(pool, 3, rng("drill-2026-08-31")));
+  });
+});

apps/api/src/srs.ts

@@ -0,0 +1,551 @@
+/**
+ * SRS domain: ET dates, the work week, the interval ladder, concept
+ * temperature, deterministic sampling, and the one write path for attempts.
+ *
+ * D1 is the single source of truth. The stage names the review a problem must
+ * pass NEXT (`+3` = due 3 working days after the last clean solve). Passing
+ * advances new → +3 → +7 → retired; any failure resets to +3. A problem's
+ * FIRST-ever log enters the ladder at +3 regardless of result: a pass earns a
+ * 3-day review, a fail must be re-solved just as soon. Two rungs, 3 and 7 —
+ * the same numbers as the `work/3` and `work/7` buckets a re-solve is
+ * scaffolded into, and as the review-issue windows in
+ * apps/cli/spaced-repetition.ts.
+ *
+ * Sunday is off, structurally: every scheduled date comes out of
+ * workingDay(), which never returns one. Nothing "falls on" the rest day and
+ * gets swallowed or carried — it is simply never booked there.
+ *
+ * All dates are America/New_York calendar strings (YYYY-MM-DD); arithmetic is
+ * anchored at noon UTC so DST edges cannot shift a date. Never raw Date math.
+ */
+import scheduleJson from "../data/schedule.json";
+
+// ── dates (America/New_York) ─────────────────────────────────────
+
+export const CAMPAIGN_START = "2026-08-17";
+export const CAMPAIGN_DAYS = 56;
+
+const ET_DATE = new Intl.DateTimeFormat("en-CA", {
+  timeZone: "America/New_York",
+  dateStyle: "short",
+});
+const ET_HOUR = new Intl.DateTimeFormat("en-US", {
+  timeZone: "America/New_York",
+  hour: "numeric",
+  hourCycle: "h23",
+});
+
+/** ET calendar date of an instant. */
+export function etDate(now: Date): string {
+  return ET_DATE.format(now);
+}
+
+/** ET hour 0–23 of an instant — the DST-proof cron guard. */
+export function etHour(now: Date): number {
+  return Number(ET_HOUR.format(now));
+}
+
+/** Noon-UTC anchor: date-only arithmetic immune to DST edges. */
+function atNoon(date: string): Date {
+  return new Date(`${date}T12:00:00Z`);
+}
+
+export function addDays(date: string, days: number): string {
+  return new Date(atNoon(date).getTime() + days * 86_400_000).toISOString().slice(0, 10);
+}
+
+export function daysBetween(from: string, to: string): number {
+  return Math.round((atNoon(to).getTime() - atNoon(from).getTime()) / 86_400_000);
+}
+
+/** 0 = Sunday … 6 = Saturday. */
+export function weekdayOf(date: string): number {
+  return atNoon(date).getUTCDay();
+}
+
+// ── the work week ────────────────────────────────────────────────
+
+/**
+ * Sunday, the one day the campaign never schedules: topics run Mon–Fri
+ * (data/schedule.json), the gate is Saturday, and the review windows below are
+ * picked so a solve comes back on a working day.
+ */
+const REST_DAY = 0;
+
+export function isRestDay(date: string): boolean {
+  return weekdayOf(date) === REST_DAY;
+}
+
+/**
+ * The `offset`-th working day counting from `from` (0 = `from` itself), with
+ * Sundays skipped — the ONE place a scheduled date is minted.
+ *
+ * Work may slide later, never earlier: pulling a review back to Saturday would
+ * shorten the very interval it exists to test, so a Sunday landing becomes
+ * Monday. With the +3/+7 ladder that only ever happens to a Thursday solve's
+ * 3-day review; +7 lands on the solve's own weekday, which is never a Sunday.
+ */
+export function workingDay(from: string, offset = 0): string {
+  let cursor = isRestDay(from) ? addDays(from, 1) : from;
+  for (let i = 0; i < offset; i++) {
+    cursor = addDays(cursor, 1);
+    if (isRestDay(cursor)) cursor = addDays(cursor, 1);
+  }
+  return cursor;
+}
+
+export function campaignDay(date: string): number {
+  return daysBetween(CAMPAIGN_START, date) + 1;
+}
+
+/** 1-based campaign week (Mon–Sun), aligned to CAMPAIGN_START. */
+export function campaignWeek(date: string): number {
+  return Math.floor(daysBetween(CAMPAIGN_START, date) / 7) + 1;
+}
+
+/** Monday of the date's campaign week. */
+export function weekMonday(date: string): string {
+  return addDays(CAMPAIGN_START, (campaignWeek(date) - 1) * 7);
+}
+
+export function isoWeek(date: string): number {
+  const d = atNoon(date);
+  d.setUTCDate(d.getUTCDate() + 4 - (d.getUTCDay() || 7));
+  const jan1 = new Date(Date.UTC(d.getUTCFullYear(), 0, 1));
+  return Math.ceil(((d.getTime() - jan1.getTime()) / 86_400_000 + 1) / 7);
+}
+
+export function prettyDate(date: string): string {
+  return new Intl.DateTimeFormat("en-US", {
+    timeZone: "UTC",
+    weekday: "long",
+    month: "long",
+    day: "numeric",
+  }).format(atNoon(date));
+}
+
+// ── the schedule (bundled; humans edit apps/api/data/schedule.json) ───
+
+export const SCHEDULE: Record<string, number> = scheduleJson;
+
+/** Date the schedule teaches a topic on, if it teaches it at all. */
+export function topicDate(topic: number): string | undefined {
+  for (const [date, t] of Object.entries(SCHEDULE)) {
+    if (t === topic) return date;
+  }
+  return undefined;
+}
+
+/** Week in which a topic was (or will be) taught. */
+export function topicWeek(topic: number): number | undefined {
+  const date = topicDate(topic);
+  return date === undefined ? undefined : campaignWeek(date);
+}
+
+// ── deterministic sampling ───────────────────────────────────────
+
+/** FNV-1a → mulberry32: seeded PRNG so re-runs pick identical problems. */
+export function rng(seed: string): () => number {
+  let h = 0x811c9dc5;
+  for (let i = 0; i < seed.length; i++) {
+    h ^= seed.charCodeAt(i);
+    h = Math.imul(h, 0x01000193);
+  }
+  return () => {
+    h = Math.imul(h ^ (h >>> 15), h | 1);
+    h ^= h + Math.imul(h ^ (h >>> 7), h | 61);
+    return ((h ^ (h >>> 14)) >>> 0) / 4294967296;
+  };
+}
+
+/** Up to n elements, Fisher–Yates order driven by the seeded PRNG. */
+export function sample<T>(pool: T[], n: number, random: () => number): T[] {
+  const copy = [...pool];
+  for (let i = copy.length - 1; i > 0; i--) {
+    const j = Math.floor(random() * (i + 1));
+    [copy[i], copy[j]] = [copy[j]!, copy[i]!];
+  }
+  return copy.slice(0, n);
+}
+
+// ── rows ─────────────────────────────────────────────────────────
+
+/**
+ * Review windows in days. One vocabulary for the whole campaign: these are the
+ * ladder rungs, the `work/<n>` buckets the picker scaffolds into, and the
+ * windows the `Spaced Repetition — <date>` issues are built from.
+ *
+ * 3 and 7 keep the rest day free. +7 returns a solve to its own weekday, and
+ * of the five learning weekdays only Thursday's +3 touches a Sunday — which
+ * workingDay() slides to Monday.
+ */
+export const WINDOWS = [3, 7] as const;
+export type Window = (typeof WINDOWS)[number];
+
+export type Stage = "new" | `+${Window}` | "retired";
+/** Chart / stats order, low to high. */
+export const STAGES: Stage[] = ["new", "+3", "+7", "retired"];
+export type Result = "pass" | "fail";
+export type Kind = "first" | "review" | "drill" | "gate";
+
+/**
+ * Catalog vocabulary, carried by the `diff:*` / `set:*` issue labels and
+ * stored verbatim in `problems`. Listed low to high / most to least required,
+ * which is the order every chart, table and digest section prints them in.
+ */
+export const DIFFICULTIES = ["easy", "medium", "hard"] as const;
+export type Difficulty = (typeof DIFFICULTIES)[number];
+export const SETS = ["core", "optional", "deferred"] as const;
+export type SetLabel = (typeof SETS)[number];
+
+/**
+ * How warm a concept is: the age of a topic's most recent attempt, bucketed by
+ * the ladder's own windows so the two cannot drift. Touched inside the first
+ * window it is `hot`, inside the second `fresh`, inside one more full second
+ * window `fading`; older than that — or never attempted — it is `cold`.
+ *
+ * A read-side label only: nothing is scheduled off it. It answers the question
+ * the ladder cannot, because the ladder tracks problems and a topic can go
+ * quiet for a fortnight while not one of its problems comes due.
+ */
+export const TEMPERATURES = ["hot", "fresh", "fading", "cold"] as const;
+export type Temperature = (typeof TEMPERATURES)[number];
+
+/** Warmest band first, with the age it tolerates. Past the last one is cold. */
+const TEMPERATURE_MAX_AGE: [Temperature, number][] = [
+  ["hot", WINDOWS[0]],
+  ["fresh", WINDOWS[1]],
+  ["fading", WINDOWS[1] * 2],
+];
+
+/** Days since a topic's last attempt → its band; null (never) is cold. */
+export function temperatureOf(age: number | null): Temperature {
+  if (age === null) return "cold";
+  for (const [temperature, max] of TEMPERATURE_MAX_AGE) {
+    if (age <= max) return temperature;
+  }
+  return "cold";
+}
+
+export interface ProblemRow {
+  lc_number: number;
+  issue: number;
+  topic_issue: number;
+  title: string;
+  difficulty: string;
+  set_label: string;
+  stage: Stage;
+  next_review: string | null;
+  defer_until: string | null;
+}
+
+export const INTERVAL: Record<string, number> = { "+3": 3, "+7": 7 };
+const NEXT_STAGE: Record<string, Stage> = { new: "+3", "+3": "+7", "+7": "retired" };
+
+// ── queries ──────────────────────────────────────────────────────
+
+export async function getProblem(db: D1Database, lc: number): Promise<ProblemRow | null> {
+  return db.prepare("SELECT * FROM problems WHERE lc_number = ?").bind(lc).first<ProblemRow>();
+}
+
+/**
+ * What "due" means, in one place: not retired, scheduled on/before ?1, and
+ * past its deferral if it has one. Every caller that asks "what is open on
+ * this day" — the digest, the review issue, the docs queue chart — shares
+ * this clause so they cannot answer differently.
+ */
+export const DUE_WHERE = `stage != 'retired' AND next_review IS NOT NULL AND next_review <= ?1
+    AND (defer_until IS NULL OR defer_until <= ?1)`;
+
+/** Reviews due on/before `date`, oldest first — the overflow carry order. */
+export async function dueReviews(db: D1Database, date: string): Promise<ProblemRow[]> {
+  const { results } = await db
+    .prepare(`SELECT * FROM problems WHERE ${DUE_WHERE} ORDER BY next_review, lc_number`)
+    .bind(date)
+    .all<ProblemRow>();
+  return results;
+}
+
+/** One problem the ladder wants re-solved today. */
+export interface DueRow {
+  lc: number;
+  issue: number;
+  title: string;
+  difficulty: string;
+  /** The rung being settled: also the `work/<n>` bucket the re-solve goes in. */
+  stage: Stage;
+  bucket: number;
+  /** Scheduled day; earlier than `date` when a day was missed. */
+  due: string;
+  /** Days late (0 = due today), so a renderer can say so. */
+  late: number;
+  /** Last logged attempt, and how long ago — the "last seen" line. */
+  last_seen: string | null;
+  days_since: number | null;
+}
+
+/**
+ * Everything the ladder wants re-solved on `date`, oldest first.
+ *
+ * The read side of "which problems are open today": the same rows the digest
+ * mails, shaped for whoever renders them (the daily Spaced Repetition issue).
+ * Nothing here writes, so it is safe to call repeatedly, and there is no
+ * second schedule to drift — D1's ladder is the schedule.
+ */
+export async function dueToday(db: D1Database, date: string): Promise<{ date: string; due: DueRow[] }> {
+  const { results } = await db
+    .prepare(
+      `SELECT lc_number, issue, title, difficulty, stage, next_review,
+              (SELECT MAX(a.date) FROM attempts a WHERE a.lc_number = problems.lc_number) AS last_seen
+         FROM problems WHERE ${DUE_WHERE} ORDER BY next_review, lc_number`,
+    )
+    .bind(date)
+    .all<{
+      lc_number: number;
+      issue: number;
+      title: string;
+      difficulty: string;
+      stage: Stage;
+      next_review: string;
+      last_seen: string | null;
+    }>();
+  return {
+    date,
+    due: results.map((r) => ({
+      lc: r.lc_number,
+      issue: r.issue,
+      title: r.title,
+      difficulty: r.difficulty,
+      stage: r.stage,
+      // "+3" -> 3: the rung's number IS its work/ bucket.
+      bucket: Number(r.stage.slice(1)),
+      due: r.next_review,
+      late: daysBetween(r.next_review, date),
+      last_seen: r.last_seen,
+      days_since: r.last_seen === null ? null : daysBetween(r.last_seen, date),
+    })),
+  };
+}
+
+/** Reviews surfaced per day; the digest levels everything past this forward. */
+export const REVIEW_CAP = 3;
+
+/**
+ * Load-level an overloaded review queue: everything due beyond today's
+ * REVIEW_CAP is pushed to a concrete future WORKING day — at most REVIEW_CAP
+ * per day, oldest first — instead of piling up as "due today". Runs every
+ * digest morning, so a future day that grows past the cap (spill plus newly
+ * maturing reviews) is simply re-levelled when it arrives. Idempotent within
+ * a date: after one pass at most REVIEW_CAP problems remain due today, so a
+ * second pass moves nothing. The spill starts tomorrow, or Monday when
+ * tomorrow is the rest day.
+ */
+export async function levelReviews(db: D1Database, date: string): Promise<number> {
+  const overflow = (await dueReviews(db, date)).slice(REVIEW_CAP);
+  if (overflow.length === 0) return 0;
+  await db.batch(
+    overflow.map((p, i) =>
+      db
+        .prepare("UPDATE problems SET next_review = ? WHERE lc_number = ?")
+        .bind(workingDay(addDays(date, 1), Math.floor(i / REVIEW_CAP)), p.lc_number),
+    ),
+  );
+  return overflow.length;
+}
+
+/**
+ * Blind drills: unsolved optional problems from topics ALREADY LEARNED —
+ * scheduled in an earlier week (the current week's optional pool is reserved
+ * for Saturday's gate) AND showing learning evidence: at least one of the
+ * topic's core problems has entered the ladder. A skipped learning day never
+ * feeds drills just because its calendar week lapsed. Never repeated,
+ * seeded by date. Boosted topics contribute up to 2 extra.
+ */
+export async function pickDrills(
+  db: D1Database,
+  date: string,
+  budget: number,
+): Promise<ProblemRow[]> {
+  if (budget <= 0) return [];
+  const week = campaignWeek(date);
+  const { results } = await db
+    .prepare(
+      `SELECT p.*, t.boost AS boost FROM problems p
+       JOIN topics t ON t.issue = p.topic_issue
+       WHERE p.set_label = 'optional' AND p.stage = 'new'
+         AND p.lc_number NOT IN (SELECT lc_number FROM drill_pool_used)
+         AND NOT EXISTS (SELECT 1 FROM attempts a WHERE a.lc_number = p.lc_number)
+         AND EXISTS (SELECT 1 FROM problems c
+                     WHERE c.topic_issue = p.topic_issue
+                       AND c.set_label = 'core' AND c.stage != 'new')
+       ORDER BY p.lc_number`,
+    )
+    .all<ProblemRow & { boost: number }>();
+  const pool = results.filter((p) => {
+    const w = topicWeek(p.topic_issue);
+    return w !== undefined && w < week;
+  });
+  const boosted = pool.filter((p) => p.boost === 1);
+  const regular = pool.filter((p) => p.boost !== 1);
+  const boostPicks = sample(boosted, Math.min(2, budget), rng(`boost-${date}`));
+  const regularPicks = sample(
+    regular,
+    Math.min(2, Math.max(0, budget - boostPicks.length)),
+    rng(`drill-${date}`),
+  );
+  return [...boostPicks, ...regularPicks].slice(0, budget);
+}
+
+/** Consecutive days with ≥1 attempt, ending today or yesterday. */
+export async function streak(db: D1Database, today: string): Promise<number> {
+  const { results } = await db
+    .prepare("SELECT DISTINCT date FROM attempts ORDER BY date DESC LIMIT 90")
+    .all<{ date: string }>();
+  const days = new Set(results.map((r) => r.date));
+  let cursor = days.has(today) ? today : addDays(today, -1);
+  let n = 0;
+  while (days.has(cursor)) {
+    n++;
+    cursor = addDays(cursor, -1);
+  }
+  return n;
+}
+
+// ── the one write path ───────────────────────────────────────────
+
+export interface LogOutcome {
+  lc: number;
+  title: string;
+  kind: Kind;
+  result: Result;
+  stage: Stage;
+  next_review: string | null;
+  /** First-ever attempt — caller closes the sub-issue on pass. */
+  first: boolean;
+  issue: number;
+  /** Email one-tap replay: nothing changed. */
+  duplicate: boolean;
+  error?: string;
+}
+
+/**
+ * Record an attempt and move the ladder. Ladder semantics live here and only
+ * here — email one-taps, webhook /done lines, gate scoring, and solutions
+ * landing in work/ all converge.
+ *
+ * Email idempotency comes from the partial unique index on
+ * (lc_number, date, kind) WHERE source='email': a replayed link inserts
+ * nothing and must not touch the ladder. Webhook corrections (pass then fail
+ * on the same day) remain legal — every webhook attempt appends.
+ *
+ * Commit idempotency needs no index either. A commit names the rung the file
+ * it landed in settles — `work/1` settles `new`, `work/3` settles `+3`,
+ * `work/7` settles `+7` — and the stage check below writes only when the
+ * ladder is actually standing on that rung. So re-pushing a solution, or the
+ * reconciler re-sending its whole implemented set on every push, moves
+ * nothing: the rung it names has already been left behind.
+ */
+export async function logAttempt(
+  db: D1Database,
+  opts: {
+    lc: number;
+    date: string;
+    result: Result;
+    source: "email" | "webhook" | "commit";
+    /** Commit only: the stage this file settles. Default `new` = first solve. */
+    rung?: Stage;
+    gate?: boolean;
+  },
+): Promise<LogOutcome> {
+  const p = await getProblem(db, opts.lc);
+  const nothing: LogOutcome = {
+    lc: opts.lc,
+    title: "",
+    kind: "review",
+    result: opts.result,
+    stage: "new",
+    next_review: null,
+    first: false,
+    issue: 0,
+    duplicate: false,
+  };
+  if (!p) return { ...nothing, error: `LC ${opts.lc} is not in the curriculum` };
+  // A commit reports a file, not an event: it may only settle the rung that
+  // file's bucket IS. Any other stage means the push is old news (the ladder
+  // already moved past it) or premature — no write either way, and a retired
+  // problem can never be woken. Fresh reviews otherwise come from the digest
+  // tap, a /done comment, or a gate.
+  if (opts.source === "commit" && p.stage !== (opts.rung ?? "new")) {
+    return {
+      ...nothing,
+      title: p.title,
+      stage: p.stage,
+      next_review: p.next_review,
+      issue: p.issue,
+      duplicate: true,
+    };
+  }
+  if (p.stage === "retired") {
+    return { ...nothing, title: p.title, issue: p.issue, error: `LC ${opts.lc} is already retired` };
+  }
+
+  const attempted = await db
+    .prepare("SELECT 1 AS x FROM attempts WHERE lc_number = ? LIMIT 1")
+    .bind(opts.lc)
+    .first();
+  const first = !attempted && p.stage === "new";
+  const kind: Kind = opts.gate ? "gate" : first ? (p.set_label === "optional" ? "drill" : "first") : "review";
+
+  const inserted = await db
+    .prepare(
+      `INSERT INTO attempts (lc_number, date, kind, result, source) VALUES (?, ?, ?, ?, ?)
+       ON CONFLICT DO NOTHING`,
+    )
+    .bind(opts.lc, opts.date, kind, opts.result, opts.source)
+    .run();
+  if (opts.source === "email" && inserted.meta.changes === 0) {
+    return { ...nothing, title: p.title, kind, stage: p.stage, next_review: p.next_review, issue: p.issue, duplicate: true };
+  }
+
+  // Ladder move. First-ever logs enter at +3 for pass AND fail.
+  let stage: Stage;
+  if (first) {
+    stage = "+3";
+  } else if (opts.result === "pass") {
+    stage = NEXT_STAGE[p.stage] ?? "retired";
+  } else {
+    stage = "+3";
+  }
+  // workingDay(), not addDays(): a Thursday solve's +3 would land on the rest
+  // day, and it takes Monday instead.
+  const next = stage === "retired" ? null : workingDay(addDays(opts.date, INTERVAL[stage]!));
+
+  const writes = [
+    db.prepare(
+      "UPDATE problems SET stage = ?, next_review = ?, defer_until = NULL WHERE lc_number = ?",
+    ).bind(stage, next, opts.lc),
+  ];
+  if (first && kind === "drill") {
+    writes.push(
+      db.prepare("INSERT OR IGNORE INTO drill_pool_used (lc_number) VALUES (?)").bind(opts.lc),
+    );
+    if (opts.result === "fail") {
+      writes.push(
+        db.prepare("UPDATE topics SET misses = misses + 1 WHERE issue = ?").bind(p.topic_issue),
+      );
+    }
+  }
+  await db.batch(writes);
+
+  return {
+    lc: opts.lc,
+    title: p.title,
+    kind,
+    result: opts.result,
+    stage,
+    next_review: next,
+    first,
+    issue: p.issue,
+    duplicate: false,
+  };
+}

apps/api/src/stats.ts

@@ -0,0 +1,230 @@
+/**
+ * GET /api/stats — the one JSON document behind the docs progress page and the
+ * `bun run stats` terminal dashboard. Read-only aggregation over D1; the shape
+ * is both readers' contract, so change all three together. A problem counts as
+ * "done" once it entered the ladder (stage != 'new').
+ *
+ * The temperature block is the one figure here that is not a count of rows:
+ * srs.ts turns a topic's days-since-last-attempt into a band, so "which
+ * concepts have gone cold" has exactly one definition and the dashboard is
+ * free to colour it however it likes.
+ */
+import {
+  CAMPAIGN_DAYS,
+  CAMPAIGN_START,
+  DIFFICULTIES,
+  DUE_WHERE,
+  addDays,
+  campaignDay,
+  campaignWeek,
+  daysBetween,
+  isoWeek,
+  STAGES,
+  streak,
+  temperatureOf,
+  topicDate,
+  type Temperature,
+} from "./srs.ts";
+
+const PHASE_NAMES: Record<number, string> = {
+  1: "I — Linear",
+  2: "II — Nodal & Grid",
+  3: "III — Hierarchical",
+  4: "IV — Relational",
+  5: "V — Decision Space",
+};
+
+/** One concept: catalog coverage plus how long since it was last exercised. */
+interface TopicStat {
+  issue: number;
+  name: string;
+  milestone: number | null;
+  /** Campaign week the schedule teaches it in; null when unscheduled. */
+  week: number | null;
+  /** False while the schedule has not reached it — temperature is moot. */
+  taught: boolean;
+  total: number;
+  done: number;
+  retired: number;
+  due: number;
+  /** ET date of the most recent attempt on any of its problems. */
+  last: string | null;
+  /** Days since `last`; null when never attempted. */
+  age: number | null;
+  temperature: Temperature;
+}
+
+async function buildTopics(db: D1Database, today: string): Promise<TopicStat[]> {
+  // Three shapes, three statements: batch() carries one row type, and casting
+  // its results apart would cost more than the round trips it saves.
+  const { results: coverage } = await db
+    .prepare(
+      `SELECT t.issue AS issue, t.name AS name, t.milestone AS milestone,
+              COUNT(p.lc_number) AS total,
+              SUM(CASE WHEN p.stage != 'new' THEN 1 ELSE 0 END) AS done,
+              SUM(CASE WHEN p.stage = 'retired' THEN 1 ELSE 0 END) AS retired
+       FROM topics t LEFT JOIN problems p ON p.topic_issue = t.issue
+       GROUP BY t.issue
+       ORDER BY t.issue`,
+    )
+    .all<{
+      issue: number;
+      name: string;
+      milestone: number | null;
+      total: number;
+      done: number | null;
+      retired: number | null;
+    }>();
+
+  const { results: activity } = await db
+    .prepare(
+      `SELECT p.topic_issue AS issue, MAX(a.date) AS last
+       FROM attempts a JOIN problems p ON p.lc_number = a.lc_number
+       GROUP BY p.topic_issue`,
+    )
+    .all<{ issue: number; last: string | null }>();
+  const lastByTopic = new Map<number, string>();
+  for (const row of activity) {
+    if (row.last) lastByTopic.set(row.issue, row.last);
+  }
+
+  const { results: dueRows } = await db
+    .prepare(
+      `SELECT topic_issue AS issue, COUNT(*) AS n
+       FROM problems WHERE ${DUE_WHERE} GROUP BY topic_issue`,
+    )
+    .bind(today)
+    .all<{ issue: number; n: number }>();
+  const dueByTopic = new Map<number, number>();
+  for (const row of dueRows) dueByTopic.set(row.issue, row.n);
+
+  return coverage.map((row) => {
+    const taughtOn = topicDate(row.issue);
+    const last = lastByTopic.get(row.issue) ?? null;
+    const age = last === null ? null : daysBetween(last, today);
+    return {
+      issue: row.issue,
+      name: row.name,
+      milestone: row.milestone,
+      week: taughtOn === undefined ? null : campaignWeek(taughtOn),
+      taught: taughtOn !== undefined && taughtOn <= today,
+      total: row.total,
+      done: row.done ?? 0,
+      retired: row.retired ?? 0,
+      due: dueByTopic.get(row.issue) ?? 0,
+      last,
+      age,
+      temperature: temperatureOf(age),
+    };
+  });
+}
+
+export async function buildStats(db: D1Database, today: string): Promise<object> {
+  const { results: phaseRows } = await db
+    .prepare(
+      `SELECT t.milestone AS milestone, p.set_label AS set_label,
+              COUNT(*) AS total,
+              SUM(CASE WHEN p.stage != 'new' THEN 1 ELSE 0 END) AS done
+       FROM problems p JOIN topics t ON t.issue = p.topic_issue
+       WHERE t.milestone IS NOT NULL
+       GROUP BY t.milestone, p.set_label
+       ORDER BY t.milestone`,
+    )
+    .all<{ milestone: number; set_label: string; total: number; done: number }>();
+  const phases = new Map<
+    number,
+    { milestone: number; name: string } & Record<string, number | string>
+  >();
+  for (const row of phaseRows) {
+    const phase =
+      phases.get(row.milestone) ??
+      ({
+        milestone: row.milestone,
+        name: PHASE_NAMES[row.milestone] ?? `Phase ${row.milestone}`,
+        core_done: 0,
+        core_total: 0,
+        optional_done: 0,
+        optional_total: 0,
+        deferred_done: 0,
+        deferred_total: 0,
+      } as { milestone: number; name: string } & Record<string, number | string>);
+    phase[`${row.set_label}_done`] = row.done;
+    phase[`${row.set_label}_total`] = row.total;
+    phases.set(row.milestone, phase);
+  }
+
+  const { results: ladderRows } = await db
+    .prepare("SELECT stage, COUNT(*) AS n FROM problems GROUP BY stage")
+    .all<{ stage: string; n: number }>();
+  // Every rung is present even at zero: the page's bar list is this object.
+  const ladder: Record<string, number> = Object.fromEntries(STAGES.map((s) => [s, 0]));
+  for (const row of ladderRows) ladder[row.stage] = row.n;
+
+  // Same shape as `ladder`: every band present, zeros included, in DIFFICULTIES
+  // order so a reader can index it without sorting.
+  const { results: diffRows } = await db
+    .prepare(
+      `SELECT difficulty,
+              COUNT(*) AS total,
+              SUM(CASE WHEN stage != 'new' THEN 1 ELSE 0 END) AS done
+       FROM problems GROUP BY difficulty`,
+    )
+    .all<{ difficulty: string; total: number; done: number }>();
+  const difficulty = DIFFICULTIES.map((name) => {
+    const row = diffRows.find((r) => r.difficulty === name);
+    return { difficulty: name, total: row?.total ?? 0, done: row?.done ?? 0 };
+  });
+
+  // gates rows are keyed by ISO week (that's the digest/gate contract), but
+  // the page speaks campaign weeks. The campaign never crosses a year
+  // boundary, so a plain offset converts safely.
+  const isoWeekOffset = isoWeek(CAMPAIGN_START) - 1;
+  const { results: gateRows } = await db
+    .prepare("SELECT week, issue, pass_rate, closed_on FROM gates ORDER BY week")
+    .all<{ week: number; issue: number | null; pass_rate: number | null; closed_on: string | null }>();
+  const gates = gateRows.map((g) => ({ ...g, week: g.week - isoWeekOffset }));
+
+  // Review-queue depth for the next 14 days: everything due by that day. One
+  // statement per day, but DUE_WHERE stays the only definition of "due" — the
+  // batch is a single round trip, not fourteen.
+  const dates = Array.from({ length: 14 }, (_, i) => addDays(today, i));
+  const queueRows = await db.batch<{ n: number }>(
+    dates.map((date) =>
+      db.prepare(`SELECT COUNT(*) AS n FROM problems WHERE ${DUE_WHERE}`).bind(date),
+    ),
+  );
+  const queue = dates.map((date, i) => ({ date, due: queueRows[i]?.results[0]?.n ?? 0 }));
+
+  // Attempts per campaign day, every day present. This is the heatmap's data
+  // model — the docs page reads the tail of it for "last 7 days" rather than
+  // asking a second question that could answer differently.
+  const end = addDays(CAMPAIGN_START, CAMPAIGN_DAYS - 1);
+  const { results: heatRows } = await db
+    .prepare(
+      `SELECT date,
+              COUNT(*) AS attempts,
+              SUM(CASE WHEN result = 'pass' THEN 1 ELSE 0 END) AS passes
+       FROM attempts WHERE date BETWEEN ?1 AND ?2 GROUP BY date`,
+    )
+    .bind(CAMPAIGN_START, end)
+    .all<{ date: string; attempts: number; passes: number | null }>();
+  const byDate = new Map(heatRows.map((r) => [r.date, r]));
+  const heat = Array.from({ length: CAMPAIGN_DAYS }, (_, i) => {
+    const date = addDays(CAMPAIGN_START, i);
+    const row = byDate.get(date);
+    return { date, attempts: row?.attempts ?? 0, passes: row?.passes ?? 0 };
+  });
+
+  return {
+    generated: new Date().toISOString(),
+    campaign: { day: campaignDay(today), week: campaignWeek(today), start: CAMPAIGN_START, days: CAMPAIGN_DAYS },
+    phases: [...phases.values()],
+    ladder,
+    difficulty,
+    topics: await buildTopics(db, today),
+    gates,
+    streak: await streak(db, today),
+    queue,
+    heat,
+  };
+}

apps/api/tsconfig.json

@@ -0,0 +1,18 @@
+{
+  "compilerOptions": {
+    "target": "esnext",
+    "module": "esnext",
+    "moduleResolution": "bundler",
+    "lib": ["esnext"],
+    "jsx": "react-jsx",
+    "strict": true,
+    "noUncheckedIndexedAccess": true,
+    "resolveJsonModule": true,
+    "allowImportingTsExtensions": true,
+    "verbatimModuleSyntax": true,
+    "skipLibCheck": true,
+    "noEmit": true,
+    "types": ["./worker-configuration.d.ts", "bun"]
+  },
+  "include": ["src", "scripts"]
+}

apps/api/wrangler.jsonc

@@ -0,0 +1,41 @@
+// SRS Worker — daily digest email, one-tap logging, weekly review issues,
+// live SVG charts, and the /api/stats feed for the docs progress page.
+//
+// D1 is the source of truth for SRS state; GitHub issues own the catalog;
+// apps/api/data/schedule.json (bundled) owns the schedule. JSONC over TOML per
+// Workers best practices (newer features are JSON-only).
+{
+  "$schema": "node_modules/wrangler/config-schema.json",
+  "name": "srs-api",
+  "main": "src/index.ts",
+  "compatibility_date": "2026-08-24",
+  "observability": { "enabled": true },
+
+  "d1_databases": [
+    { "binding": "DB", "database_name": "srs", "database_id": "2ee2c34a-c2e2-4ebb-934e-c9e7dc8aa4f1" }
+  ],
+
+  // Sender domain prdlk.com is onboarded to Email Service; the binding is
+  // restricted to the one verified destination inbox.
+  "send_email": [
+    { "name": "EMAIL", "allowed_destination_addresses": ["prnk28@gmail.com"] }
+  ],
+
+  // Double-cron per event makes each DST-proof: the code fires only when the
+  // computed ET hour matches (8 AM daily digest, midnight-Saturday review).
+  // 8 AM ET = 12:00 UTC in EDT, 13:00 UTC in EST; midnight ET = 4/5 UTC.
+  "triggers": {
+    "crons": ["0 12 * * *", "0 13 * * *", "0 4 * * 6", "0 5 * * 6"]
+  },
+
+  "vars": {
+    "REPO": "prdlk/leetcode",
+    "FROM_EMAIL": "srs@prdlk.com",
+    "TO_EMAIL": "prnk28@gmail.com",
+    "DOCS_URL": "https://prdlk.github.io/leetcode",
+    "DOCS_ORIGIN": "https://prdlk.github.io,https://lc.prad.nu",
+    "PUBLIC_URL": "https://srs-api.prdlk.workers.dev"
+  }
+
+  // Secrets (wrangler secret put): GH_PAT, WEBHOOK_SECRET, LINK_KEY.
+}

apps/cli/api.ts

@@ -0,0 +1,87 @@
+/**
+ * The SRS Worker, from the CLI side: one base URL and one read-side payload.
+ *
+ * `SRS_API` points a script at a `wrangler dev` instance instead of production
+ * — the same switch close-solved.ts uses for its `/admin/solved` push, which is
+ * why the default lives here rather than in either script.
+ *
+ * The payload is narrowed with a guard rather than cast: D1 is the only source
+ * of SRS truth, so if the Worker answers something else the caller must say so
+ * out loud, not render half a dashboard off undefined fields.
+ */
+import { STAGES, TEMPERATURES, type Stage, type Temperature } from "../api/src/srs.ts";
+
+export const SRS_API = process.env.SRS_API ?? "https://srs-api.prdlk.workers.dev";
+
+// The rungs and the bands come straight from the Worker's domain module — the
+// same import work.ts makes for WINDOWS — so a change to either vocabulary
+// reaches this dashboard without a second list to remember.
+export { STAGES, TEMPERATURES, type Stage, type Temperature };
+
+export interface Phase {
+  milestone: number;
+  name: string;
+  core_done: number;
+  core_total: number;
+  optional_done: number;
+  optional_total: number;
+  deferred_done: number;
+  deferred_total: number;
+}
+
+export interface Topic {
+  issue: number;
+  name: string;
+  milestone: number | null;
+  week: number | null;
+  taught: boolean;
+  total: number;
+  done: number;
+  retired: number;
+  due: number;
+  last: string | null;
+  age: number | null;
+  temperature: Temperature;
+}
+
+export interface Stats {
+  generated: string;
+  campaign: { day: number; week: number; start: string; days: number };
+  phases: Phase[];
+  ladder: Record<Stage, number>;
+  difficulty: { difficulty: string; total: number; done: number }[];
+  topics: Topic[];
+  /** week is the 1-based campaign week — the Worker converts from ISO. */
+  gates: { week: number; issue: number | null; pass_rate: number | null; closed_on: string | null }[];
+  streak: number;
+  queue: { date: string; due: number }[];
+  /** Every campaign day, zeros and future days included. */
+  heat: { date: string; attempts: number; passes: number }[];
+}
+
+function isStats(value: unknown): value is Stats {
+  if (typeof value !== "object" || value === null) return false;
+  const v = value as Record<string, unknown>;
+  return (
+    typeof v.campaign === "object" &&
+    v.campaign !== null &&
+    typeof v.ladder === "object" &&
+    v.ladder !== null &&
+    typeof v.streak === "number" &&
+    Array.isArray(v.phases) &&
+    Array.isArray(v.difficulty) &&
+    Array.isArray(v.topics) &&
+    Array.isArray(v.gates) &&
+    Array.isArray(v.queue) &&
+    Array.isArray(v.heat)
+  );
+}
+
+/** GET /api/stats. Throws with the reason; there is no local substitute. */
+export async function fetchStats(signal?: AbortSignal): Promise<Stats> {
+  const res = await fetch(`${SRS_API}/api/stats`, { signal });
+  if (!res.ok) throw new Error(`GET /api/stats -> ${res.status} ${await res.text()}`);
+  const body: unknown = await res.json();
+  if (!isStats(body)) throw new Error("GET /api/stats -> unrecognised payload shape");
+  return body;
+}

apps/cli/close-solved.ts

@@ -0,0 +1,232 @@
+#!/usr/bin/env bun
+/**
+ * Close the GitHub issue for every LeetCode problem actually solved under work/,
+ * then report the solved set to the SRS Worker so D1 agrees.
+ *
+ * Reconciles state instead of reacting to a push diff: any `problem`-labelled
+ * issue whose LC number has an *implemented* solution file in work/ gets
+ * closed. Only open issues are touched, so re-runs are no-ops and backfilling
+ * needs no special casing. Removing a solution never reopens an issue.
+ *
+ * `bun run pick` scaffolds a statement header plus an empty function body, so
+ * file existence alone means nothing — a stub must not close its issue. See
+ * isImplemented() in source.ts.
+ *
+ *   bun apps/cli/close-solved.ts            # close matches, log them in D1
+ *   bun apps/cli/close-solved.ts --dry-run  # report only, touch nothing
+ *
+ * Auth: GH_TOKEN / GITHUB_TOKEN, else falls back to `gh auth token`.
+ *       SRS_ADMIN_KEY for the D1 push (unset = skip it); SRS_API overrides
+ *       the Worker URL for local `wrangler dev` runs.
+ */
+import { join } from "node:path";
+
+import { SRS_API } from "./api.ts";
+import { github } from "./github.ts";
+import { markdownTable, printTable, writeStepSummary } from "./report.ts";
+import { isImplemented } from "./source.ts";
+import { WORK, parseWorkPath, type Bucket } from "./work.ts";
+
+const DRY =
+  process.argv.includes("--dry-run") ||
+  process.env.DRY_RUN === "1" ||
+  process.env.DRY_RUN === "true";
+
+// ── repo + auth ──────────────────────────────────────────────────
+
+const gh = await github();
+
+// ── work/ inventory ──────────────────────────────────────────────
+
+interface WorkEntry {
+  files: string[];
+  /** Any bucket implemented — this is what closes the problem issue. */
+  implemented: boolean;
+  /** Buckets holding a real solution: 1 = first solve, 3/7 = the re-solves. */
+  solved: Set<Bucket>;
+}
+
+/** LC number -> its solution files across every bucket (js and py both count). */
+const work = new Map<number, WorkEntry>();
+for await (const rel of new Bun.Glob("**/*.{js,py}").scan(WORK)) {
+  const file = parseWorkPath(rel);
+  if (!file) {
+    console.warn(`skip (off-layout): work/${rel}`);
+    continue;
+  }
+  const src = await Bun.file(join(WORK, rel)).text();
+  const implemented = isImplemented(src, rel.endsWith(".py") ? "py" : "js");
+  const entry = work.get(file.lc) ?? { files: [], implemented: false, solved: new Set<Bucket>() };
+  entry.files.push(`work/${rel}`);
+  entry.implemented ||= implemented;
+  if (implemented) entry.solved.add(file.bucket);
+  work.set(file.lc, entry);
+}
+
+// ── open problem issues ──────────────────────────────────────────
+
+interface ProblemIssue {
+  number: number;
+  set: string;
+}
+
+/**
+ * Narrow one element of the /issues payload to the fields this script needs.
+ * Returns undefined for pull requests and for anything whose title is not a
+ * `LC <num> ...` problem, which is how non-curriculum rows get skipped.
+ */
+function readProblemIssue(
+  value: unknown,
+): { lc: number; issue: ProblemIssue } | undefined {
+  if (!value || typeof value !== "object") return;
+  if ("pull_request" in value) return; // the /issues route also lists PRs
+  if (!("number" in value) || typeof value.number !== "number") return;
+  if (!("title" in value) || typeof value.title !== "string") return;
+  const lc = value.title.match(/^LC (\d+) /);
+  if (!lc) return;
+
+  let set = "—";
+  if ("labels" in value && Array.isArray(value.labels)) {
+    for (const label of value.labels) {
+      if (
+        label &&
+        typeof label === "object" &&
+        "name" in label &&
+        typeof label.name === "string" &&
+        label.name.startsWith("set:")
+      ) {
+        set = label.name;
+      }
+    }
+  }
+  return { lc: Number(lc[1]), issue: { number: value.number, set } };
+}
+
+/** LC number -> open issue carrying the `problem` label. */
+const open = new Map<number, ProblemIssue>();
+for await (const raw of gh.list(`/repos/${gh.repo}/issues?labels=problem&state=open`)) {
+  const parsed = readProblemIssue(raw);
+  if (parsed) open.set(parsed.lc, parsed.issue);
+}
+
+// ── reconcile ────────────────────────────────────────────────────
+
+const sha = process.env.GITHUB_SHA;
+const report: string[][] = [];
+let closed = 0;
+
+for (const num of [...work.keys()].sort((a, b) => a - b)) {
+  const entry = work.get(num)!;
+  const issue = open.get(num);
+
+  // Warm-ups and out-of-curriculum practice have no issue. Not an error.
+  if (!issue) {
+    report.push([`${num}`, "—", "—", "no open issue"]);
+    continue;
+  }
+  if (!entry.implemented) {
+    report.push([`${num}`, `#${issue.number}`, issue.set, "stub — skipped"]);
+    continue;
+  }
+  if (DRY) {
+    report.push([`${num}`, `#${issue.number}`, issue.set, "would close"]);
+    continue;
+  }
+
+  const links = entry.files
+    .map((f) =>
+      sha ? `[\`${f}\`](https://github.com/${gh.repo}/blob/${sha}/${encodeURI(f)})` : `\`${f}\``,
+    )
+    .join(", ");
+  await gh.closeIssue(
+    issue.number,
+    `Solved — solution committed at ${links}.\n\n` +
+      "Closed automatically by `close-solved`. Fill in the close-out block above if you have not already.",
+  );
+  report.push([`${num}`, `#${issue.number}`, issue.set, "closed"]);
+  closed++;
+}
+
+// ── push the solved set into D1 ───────────────────────────────────
+
+// Closing an issue is invisible to the Worker: its catalog reconcile reads
+// GitHub for titles, labels and milestones and never looks at issue state,
+// and D1's SRS columns only ever move through logAttempt(). So a solution
+// pushed here — rather than tapped in the digest email — has to be reported,
+// or the charts, the digest's "already solved" ticks, and the drill/gate
+// pools all keep treating it as untouched.
+//
+// Each entry carries the BUCKET the file sits in, which is the rung it
+// settles: work/1 is the first solve, work/3 the 3-day review, work/7 the
+// 7-day one. That is what makes a pushed re-solve advance the ladder instead
+// of vanishing — and the Worker writes only when a problem is standing on the
+// rung named, so the whole implemented set can go over on every push (a total
+// recompute that backfills whatever earlier runs missed) and re-runs write
+// nothing.
+// SRS_API (apps/cli/api.ts) is the switch that points this at a wrangler dev
+// instance instead of production.
+const solved = [...work.entries()]
+  .flatMap(([lc, entry]) => [...entry.solved].sort().map((bucket) => ({ lc, bucket })))
+  .sort((a, b) => a.lc - b.lc || a.bucket - b.bucket);
+
+let syncNote: string;
+let syncFailed = false;
+if (!process.env.SRS_ADMIN_KEY) {
+  syncNote = "SRS_ADMIN_KEY unset — D1 not touched";
+} else if (solved.length === 0) {
+  syncNote = "nothing implemented — D1 not touched";
+} else {
+  try {
+    const res = await fetch(`${SRS_API}/admin/solved${DRY ? "?dry=1" : ""}`, {
+      method: "POST",
+      headers: {
+        authorization: `Bearer ${process.env.SRS_ADMIN_KEY}`,
+        "content-type": "application/json",
+      },
+      body: JSON.stringify({ solved }),
+    });
+    if (!res.ok) throw new Error(`${res.status} ${await res.text()}`);
+    // Narrow the report payload instead of casting it; a shape change should
+    // read as "logged nothing", not throw inside the reconciler.
+    const payload: unknown = await res.json();
+    const fields = payload && typeof payload === "object" ? payload : {};
+    const strings = (value: unknown): string[] => (Array.isArray(value) ? value.map(String) : []);
+    const logged = strings("logged" in fields ? fields.logged : null);
+    const skipped = strings("skipped" in fields ? fields.skipped : null);
+    syncNote =
+      `${DRY ? "would log" : "logged"} ${logged.length} of ${solved.length} ` +
+      `(${skipped.length} already on the ladder or off-curriculum)`;
+    for (const line of logged) console.log(`  d1: ${line}`);
+  } catch (err) {
+    // The closes above already landed; surface the failure instead of
+    // letting D1 drift silently until the next push.
+    syncFailed = true;
+    syncNote = `D1 push FAILED: ${err instanceof Error ? err.message : String(err)}`;
+  }
+}
+
+// ── report ───────────────────────────────────────────────────────
+
+const rows = [["lc", "issue", "set", "status"], ...report];
+printTable(rows);
+
+const actionable = report.filter((r) => r[3] !== "no open issue");
+console.log(
+  `\n${work.size} problems in work/ · ${solved.length} solved files across buckets · ` +
+    `${actionable.length} matched an open issue · ` +
+    `${DRY ? `would close ${actionable.filter((r) => r[3] === "would close").length}` : `closed ${closed}`}` +
+    `\nd1: ${syncNote}`,
+);
+
+await writeStepSummary(
+  `### close-solved${DRY ? " (dry run)" : ""}\n\n` +
+    `${work.size} problems in \`work/\`, ${solved.length} solved files across buckets, ` +
+    `${actionable.length} matched an open issue.\n\n` +
+    (actionable.length
+      ? `${markdownTable([rows[0]!, ...actionable])}\n`
+      : "Nothing to close.\n") +
+    `\nD1: ${syncNote}\n`,
+);
+
+// Non-zero only for the D1 push: the closes are already reported above.
+if (syncFailed) process.exitCode = 1;

apps/cli/close-topics.ts

@@ -0,0 +1,150 @@
+#!/usr/bin/env bun
+/**
+ * Close the GitHub issue for every topic whose required (core) problems are done.
+ *
+ * Each `topic`-labelled issue owns its problems as GitHub sub-issues, and each
+ * problem carries exactly one `set:` label — `set:core` is required, while
+ * `set:optional` and `set:deferred` are extra credit. A topic is finished when
+ * every one of its core sub-issues is closed; optional/deferred state is
+ * ignored, matching the day rule in each topic body ("Core problems first").
+ *
+ * Like close-solved.ts this reconciles state instead of reacting to an event
+ * payload: re-runs are no-ops, backfilling needs no special casing, and only
+ * open topics are touched, so reopening a problem never reopens its topic.
+ *
+ *   bun apps/cli/close-topics.ts            # close finished topics
+ *   bun apps/cli/close-topics.ts --dry-run  # report only, touch nothing
+ *
+ * Auth: GH_TOKEN / GITHUB_TOKEN, else falls back to `gh auth token`.
+ */
+import { github } from "./github.ts";
+import { markdownTable, printTable, writeStepSummary } from "./report.ts";
+
+const DRY =
+  process.argv.includes("--dry-run") ||
+  process.env.DRY_RUN === "1" ||
+  process.env.DRY_RUN === "true";
+
+const gh = await github();
+
+// ── open topic issues ────────────────────────────────────────────
+
+interface Topic {
+  number: number;
+  title: string;
+}
+
+/**
+ * Narrow one element of the /issues payload to the fields this script needs.
+ * The `topic` label filter cannot exclude pull requests, so drop those here.
+ */
+function readTopic(value: unknown): Topic | undefined {
+  if (!value || typeof value !== "object") return;
+  if ("pull_request" in value) return; // the /issues route also lists PRs
+  if (!("number" in value) || typeof value.number !== "number") return;
+  if (!("title" in value) || typeof value.title !== "string") return;
+  return { number: value.number, title: value.title };
+}
+
+const topics: Topic[] = [];
+for await (const raw of gh.list(`/repos/${gh.repo}/issues?labels=topic&state=open`)) {
+  const topic = readTopic(raw);
+  if (topic) topics.push(topic);
+}
+topics.sort((a, b) => a.number - b.number);
+
+// ── core sub-issue state per topic ───────────────────────────────
+
+interface Core {
+  /** Core sub-issues, closed and open alike, lowest number first. */
+  all: number[];
+  /** The core sub-issues still open — non-empty means the topic stays open. */
+  pending: number[];
+}
+
+/**
+ * Read a topic's core sub-issue state. The sub_issues payload carries the full
+ * issue objects (state + labels), so no per-problem follow-up request is needed.
+ */
+async function readCore(topic: number): Promise<Core> {
+  const all: number[] = [];
+  const pending: number[] = [];
+  for await (const raw of gh.list(`/repos/${gh.repo}/issues/${topic}/sub_issues`)) {
+    if (!raw || typeof raw !== "object") continue;
+    if (!("number" in raw) || typeof raw.number !== "number") continue;
+    if (!("state" in raw) || typeof raw.state !== "string") continue;
+    if (!("labels" in raw) || !Array.isArray(raw.labels)) continue;
+
+    const core = raw.labels.some(
+      (label) =>
+        label &&
+        typeof label === "object" &&
+        "name" in label &&
+        label.name === "set:core",
+    );
+    if (!core) continue;
+
+    all.push(raw.number);
+    if (raw.state === "open") pending.push(raw.number);
+  }
+  all.sort((a, b) => a - b);
+  pending.sort((a, b) => a - b);
+  return { all, pending };
+}
+
+// ── reconcile ────────────────────────────────────────────────────
+
+const report: string[][] = [];
+let closed = 0;
+
+for (const topic of topics) {
+  const core = await readCore(topic.number);
+  const progress = `${core.all.length - core.pending.length}/${core.all.length}`;
+
+  // A topic with no core sub-issues has nothing to complete: never close it,
+  // since that would be indistinguishable from a mis-labelled problem set.
+  if (core.all.length === 0) {
+    report.push([`#${topic.number}`, topic.title, progress, "no core set"]);
+    continue;
+  }
+  if (core.pending.length > 0) {
+    report.push([
+      `#${topic.number}`,
+      topic.title,
+      progress,
+      `open: ${core.pending.map((n) => `#${n}`).join(" ")}`,
+    ]);
+    continue;
+  }
+  if (DRY) {
+    report.push([`#${topic.number}`, topic.title, progress, "would close"]);
+    continue;
+  }
+
+  await gh.closeIssue(
+    topic.number,
+    `Core set complete — all ${core.all.length} core problems closed ` +
+      `(${core.all.map((n) => `#${n}`).join(", ")}).\n\n` +
+      "Closed automatically by `close-topics`. Optional and deferred problems " +
+      "stay open as extra credit.",
+  );
+  report.push([`#${topic.number}`, topic.title, progress, "closed"]);
+  closed++;
+}
+
+// ── report ───────────────────────────────────────────────────────
+
+const rows = [["topic", "title", "core", "status"], ...report];
+printTable(rows);
+
+const finished = report.filter((r) => r[3] === "closed" || r[3] === "would close");
+console.log(
+  `\n${topics.length} open topics · ` +
+    `${DRY ? `would close ${finished.length}` : `closed ${closed}`}`,
+);
+
+await writeStepSummary(
+  `### close-topics${DRY ? " (dry run)" : ""}\n\n` +
+    `${topics.length} open topic issues, ${finished.length} with a complete core set.\n\n` +
+    (report.length ? `${markdownTable(rows)}\n` : "No open topics.\n"),
+);

apps/cli/db.ts

@@ -0,0 +1,235 @@
+/**
+ * Local LeetCode problem cache backed by SQLite at
+ * ~/.local/share/leetcode/data.db (XDG_DATA_HOME honored).
+ *
+ * One `problems` table, two freshness rules:
+ *   - the full problem index (id / title / slug / difficulty / paid flag) is
+ *     refreshed from leetcode.com at most once per 24h (`meta.index_fetched_at`);
+ *   - descriptions (HTML from GraphQL `question.content`) are fetched once per
+ *     slug and kept forever. Only the campaign problems linked from README.md
+ *     are ever fetched, and paid-only problems are skipped because GraphQL
+ *     returns null content for them.
+ *
+ * A `solutions` table tracks the lifecycle of each work/ file by LC number:
+ * `created_at` (backfilled from the file's birthtime on every picker run),
+ * `last_tested` and `last_submitted` (stamped on exit 0 of leetcode-cli).
+ * The test picker sorts by last_tested, the submit picker by created_at, and
+ * submit hides anything with a last_submitted.
+ *
+ * Library: side-effect-free on import — nothing is opened or fetched until
+ * problemDb() is called. Index upserts never touch the description column, so
+ * a daily refresh cannot evict cached statements.
+ */
+import { Database } from "bun:sqlite";
+import { mkdirSync } from "node:fs";
+import { homedir } from "node:os";
+import { dirname, join } from "node:path";
+
+export interface Problem {
+  id: number;
+  title: string;
+  slug: string;
+  difficulty: "Easy" | "Medium" | "Hard";
+  paidOnly: boolean;
+}
+
+/** Lifecycle of one work/ solution, keyed by LC number. ISO timestamps. */
+export interface SolutionActivity {
+  createdAt: string;
+  lastTested: string | null;
+  lastSubmitted: string | null;
+}
+
+const DIFFICULTY = ["", "Easy", "Medium", "Hard"] as const;
+const INDEX_TTL_MS = 24 * 60 * 60 * 1000;
+
+export const DB_PATH = join(
+  process.env.XDG_DATA_HOME ?? join(homedir(), ".local", "share"),
+  "leetcode",
+  "data.db",
+);
+
+// ── README parsing ───────────────────────────────────────────────
+
+/** Campaign problem slugs linked from README.md, in order, deduplicated. */
+export function readmeSlugs(markdown: string): string[] {
+  const slugs = new Set<string>();
+  for (const m of markdown.matchAll(/https:\/\/leetcode\.com\/problems\/([\w-]+)\//g)) {
+    slugs.add(m[1]!);
+  }
+  return [...slugs];
+}
+
+// ── remote fetches ───────────────────────────────────────────────
+
+async function fetchIndex(): Promise<Problem[]> {
+  const res = await fetch("https://leetcode.com/api/problems/all/", {
+    headers: { "user-agent": "Mozilla/5.0" },
+  });
+  if (!res.ok) throw new Error(`GET /api/problems/all -> ${res.status} ${res.statusText}`);
+  const data = (await res.json()) as {
+    stat_status_pairs: Array<{
+      stat: {
+        frontend_question_id: number;
+        question__title: string;
+        question__title_slug: string;
+      };
+      difficulty: { level: 1 | 2 | 3 };
+      paid_only: boolean;
+    }>;
+  };
+  return data.stat_status_pairs
+    .map((p) => ({
+      id: p.stat.frontend_question_id,
+      title: p.stat.question__title,
+      slug: p.stat.question__title_slug,
+      difficulty: DIFFICULTY[p.difficulty.level] as Problem["difficulty"],
+      paidOnly: p.paid_only,
+    }))
+    .sort((a, b) => a.id - b.id);
+}
+
+/** HTML problem statement, or null when LeetCode withholds it (paid-only). */
+export async function fetchDescription(slug: string): Promise<string | null> {
+  const res = await fetch("https://leetcode.com/graphql", {
+    method: "POST",
+    headers: {
+      "content-type": "application/json",
+      "user-agent": "Mozilla/5.0",
+      referer: `https://leetcode.com/problems/${slug}/`,
+    },
+    body: JSON.stringify({
+      query: "query questionContent($titleSlug: String!) { question(titleSlug: $titleSlug) { content } }",
+      variables: { titleSlug: slug },
+    }),
+  });
+  if (!res.ok) throw new Error(`POST /graphql (${slug}) -> ${res.status} ${res.statusText}`);
+  const data = (await res.json()) as { data?: { question?: { content?: string | null } | null } };
+  return data.data?.question?.content ?? null;
+}
+
+// ── the cache ────────────────────────────────────────────────────
+
+export function problemDb(path = DB_PATH) {
+  mkdirSync(dirname(path), { recursive: true });
+  const db = new Database(path);
+  db.exec(`
+    CREATE TABLE IF NOT EXISTS problems (
+      id          INTEGER PRIMARY KEY,
+      title       TEXT NOT NULL,
+      slug        TEXT NOT NULL UNIQUE,
+      difficulty  TEXT NOT NULL,
+      paid_only   INTEGER NOT NULL,
+      description TEXT
+    );
+    CREATE TABLE IF NOT EXISTS solutions (
+      id             INTEGER PRIMARY KEY,
+      created_at     TEXT NOT NULL,
+      last_tested    TEXT,
+      last_submitted TEXT
+    );
+    DROP TABLE IF EXISTS submissions;
+    CREATE TABLE IF NOT EXISTS meta (
+      key   TEXT PRIMARY KEY,
+      value TEXT NOT NULL
+    );
+  `);
+
+  const getMeta = db.query<{ value: string }, [string]>("SELECT value FROM meta WHERE key = ?1");
+  const setMeta = db.query<never, [string, string]>(
+    "INSERT INTO meta (key, value) VALUES (?1, ?2) ON CONFLICT(key) DO UPDATE SET value = excluded.value",
+  );
+  const upsert = db.query<never, [number, string, string, string, number]>(
+    `INSERT INTO problems (id, title, slug, difficulty, paid_only)
+     VALUES (?1, ?2, ?3, ?4, ?5)
+     ON CONFLICT(id) DO UPDATE SET
+       title = excluded.title, slug = excluded.slug,
+       difficulty = excluded.difficulty, paid_only = excluded.paid_only`,
+  );
+  const descState = db.query<{ paid_only: number; described: number }, [string]>(
+    "SELECT paid_only, description IS NOT NULL AS described FROM problems WHERE slug = ?1",
+  );
+
+  return {
+    path,
+
+    /** True when the index has never been fetched or is older than 24h. */
+    indexStale(): boolean {
+      const row = getMeta.get("index_fetched_at");
+      return !row || Date.now() - Number(row.value) > INDEX_TTL_MS;
+    },
+
+    /** Pull the full index from leetcode.com and upsert it. Returns the count. */
+    async refreshIndex(): Promise<number> {
+      const problems = await fetchIndex();
+      db.transaction(() => {
+        for (const p of problems) upsert.run(p.id, p.title, p.slug, p.difficulty, p.paidOnly ? 1 : 0);
+        setMeta.run("index_fetched_at", String(Date.now()));
+      })();
+      return problems.length;
+    },
+
+    /** Every cached problem, ordered by LC number. */
+    all(): Problem[] {
+      const rows = db
+        .query<{ id: number; title: string; slug: string; difficulty: string; paid_only: number }, []>(
+          "SELECT id, title, slug, difficulty, paid_only FROM problems ORDER BY id",
+        )
+        .all();
+      return rows.map((r) => ({
+        id: r.id,
+        title: r.title,
+        slug: r.slug,
+        difficulty: r.difficulty as Problem["difficulty"],
+        paidOnly: r.paid_only === 1,
+      }));
+    },
+
+    /** Free (non-premium) slugs among `slugs` with no cached description yet. */
+    missingDescriptions(slugs: string[]): string[] {
+      return slugs.filter((slug) => {
+        const row = descState.get(slug);
+        return row !== null && !row.paid_only && !row.described;
+      });
+    },
+
+    saveDescription(slug: string, html: string): void {
+      db.query("UPDATE problems SET description = ?2 WHERE slug = ?1").run(slug, html);
+    },
+
+    /** Per-solution lifecycle rows keyed by LC number. */
+    solutionActivity(): Map<number, SolutionActivity> {
+      const rows = db
+        .query<{ id: number; created_at: string; last_tested: string | null; last_submitted: string | null }, []>(
+          "SELECT id, created_at, last_tested, last_submitted FROM solutions",
+        )
+        .all();
+      return new Map(
+        rows.map((r) => [r.id, { createdAt: r.created_at, lastTested: r.last_tested, lastSubmitted: r.last_submitted }]),
+      );
+    },
+
+    /** Register a solution's creation time; existing rows are left alone. */
+    ensureSolution(id: number, createdAt: string): void {
+      db.query("INSERT INTO solutions (id, created_at) VALUES (?1, ?2) ON CONFLICT(id) DO NOTHING").run(id, createdAt);
+    },
+
+    markTested(id: number): void {
+      db.query(
+        `INSERT INTO solutions (id, created_at, last_tested) VALUES (?1, ?2, ?2)
+         ON CONFLICT(id) DO UPDATE SET last_tested = excluded.last_tested`,
+      ).run(id, new Date().toISOString());
+    },
+
+    markSubmitted(id: number): void {
+      db.query(
+        `INSERT INTO solutions (id, created_at, last_submitted) VALUES (?1, ?2, ?2)
+         ON CONFLICT(id) DO UPDATE SET last_submitted = excluded.last_submitted`,
+      ).run(id, new Date().toISOString());
+    },
+
+    close(): void {
+      db.close();
+    },
+  };
+}

apps/cli/github.ts

@@ -0,0 +1,113 @@
+/**
+ * GitHub REST + GraphQL plumbing shared by the issue reconcilers (close-solved,
+ * close-topics) and the Project field reconciler (sync-project-fields).
+ *
+ * Importing this module is side-effect free — nothing resolves credentials or
+ * touches the network until github() is awaited — so a script can import it
+ * without inheriting another script's pipeline.
+ *
+ * Repo:  GITHUB_REPOSITORY, else the `origin` remote.
+ * Auth:  GH_TOKEN / GITHUB_TOKEN, else `gh auth token`.
+ */
+import { $ } from "bun";
+import { join } from "node:path";
+
+const ROOT = join(import.meta.dir, "..", "..");
+
+/** Page size used for every list endpoint; also the "more pages" threshold. */
+const PER_PAGE = 100;
+
+export interface GitHub {
+  /** `owner/name`. */
+  repo: string;
+  /** One authenticated request against api.github.com; throws on non-2xx. */
+  api(path: string, init?: RequestInit): Promise<unknown>;
+  /** One GraphQL call (Projects v2 lives here); throws on transport or query errors. */
+  graphql(query: string, variables?: Record<string, unknown>): Promise<unknown>;
+  /** Every page of a list endpoint, flattened into one stream of elements. */
+  list(path: string): AsyncGenerator<unknown, void, void>;
+  /** Comment on an issue, then close it as completed. */
+  closeIssue(issue: number, comment: string): Promise<void>;
+}
+
+async function resolveRepo(): Promise<string> {
+  if (process.env.GITHUB_REPOSITORY) return process.env.GITHUB_REPOSITORY;
+  const url = (await $`git -C ${ROOT} remote get-url origin`.text()).trim();
+  const m = url.match(/github\.com[:/](.+?)(?:\.git)?$/);
+  if (!m) throw new Error(`cannot derive owner/repo from remote: ${url}`);
+  return m[1]!;
+}
+
+async function resolveToken(): Promise<string> {
+  const env = process.env.GH_TOKEN || process.env.GITHUB_TOKEN;
+  if (env) return env;
+  const token = (await $`gh auth token`.text()).trim();
+  if (!token) throw new Error("no credentials: set GH_TOKEN or run `gh auth login`");
+  return token;
+}
+
+export async function github(): Promise<GitHub> {
+  const repo = await resolveRepo();
+  const token = await resolveToken();
+
+  async function api(path: string, init: RequestInit = {}): Promise<unknown> {
+    const res = await fetch(`https://api.github.com${path}`, {
+      ...init,
+      headers: {
+        accept: "application/vnd.github+json",
+        authorization: `Bearer ${token}`,
+        "x-github-api-version": "2022-11-28",
+        ...(init.body ? { "content-type": "application/json" } : {}),
+      },
+    });
+    if (!res.ok) {
+      throw new Error(`${init.method ?? "GET"} ${path} -> ${res.status} ${await res.text()}`);
+    }
+    return res.json();
+  }
+
+  async function* list(path: string): AsyncGenerator<unknown, void, void> {
+    const sep = path.includes("?") ? "&" : "?";
+    for (let page = 1; ; page++) {
+      const batch = await api(`${path}${sep}per_page=${PER_PAGE}&page=${page}`);
+      if (!Array.isArray(batch)) throw new Error(`unexpected ${path} payload: not an array`);
+      yield* batch;
+      if (batch.length < PER_PAGE) return;
+    }
+  }
+
+  async function graphql(
+    query: string,
+    variables: Record<string, unknown> = {},
+  ): Promise<unknown> {
+    const res = await fetch("https://api.github.com/graphql", {
+      method: "POST",
+      headers: {
+        authorization: `Bearer ${token}`,
+        "content-type": "application/json",
+      },
+      body: JSON.stringify({ query, variables }),
+    });
+    if (!res.ok) throw new Error(`graphql -> ${res.status} ${await res.text()}`);
+    const payload = (await res.json()) as { data?: unknown; errors?: { message: string }[] };
+    if (payload.errors?.length) throw new Error(payload.errors.map((e) => e.message).join("; "));
+    return payload.data;
+  }
+
+  /**
+   * Comment before closing: if the PATCH fails, the issue still carries a
+   * visible note of what the automation decided, instead of failing silently.
+   */
+  async function closeIssue(issue: number, comment: string): Promise<void> {
+    await api(`/repos/${repo}/issues/${issue}/comments`, {
+      method: "POST",
+      body: JSON.stringify({ body: comment }),
+    });
+    await api(`/repos/${repo}/issues/${issue}`, {
+      method: "PATCH",
+      body: JSON.stringify({ state: "closed", state_reason: "completed" }),
+    });
+  }
+
+  return { repo, api, graphql, list, closeIssue };
+}

apps/cli/package.json

@@ -0,0 +1,19 @@
+{
+  "name": "cli",
+  "private": true,
+  "type": "module",
+  "scripts": {
+    "pick": "bun ./pick.ts",
+    "stats": "bun ./stats.ts",
+    "test": "bun ./test.ts",
+    "submit": "bun ./submit.ts",
+    "sync": "bun ./sync.ts",
+    "close-solved": "bun ./close-solved.ts",
+    "close-topics": "bun ./close-topics.ts"
+  },
+  "devDependencies": {
+    "@clack/prompts": "^1.7.0",
+    "@night-slayer18/leetcode-cli": "^3.5.0",
+    "@types/bun": "^1.3.14"
+  }
+}

apps/cli/pick.ts

@@ -0,0 +1,255 @@
+#!/usr/bin/env bun
+/**
+ * Fuzzy-pick a problem and scaffold it via leetcode-cli, in one of the three
+ * work/ buckets (see work.ts): a first solve under work/1, or the 3-day and
+ * 7-day blind re-solves under work/3 and work/7. Tab cycles the sections.
+ *
+ * The problem index lives in the local SQLite cache (see db.ts), refreshed at
+ * most once per 24h; campaign problem descriptions from README.md are
+ * backfilled into the same db on the way. The new-problem section hides
+ * anything that already has a numbered file under work/1 — re-picking it would
+ * only clobber the existing scaffold — and each review section hides whatever
+ * its own bucket already holds, so a finished re-solve leaves the list.
+ */
+import { autocomplete, cancel, isCancel, log, spinner } from "@clack/prompts";
+import { existsSync, mkdirSync, readFileSync, writeFileSync } from "node:fs";
+import { homedir } from "node:os";
+import { join } from "node:path";
+import { styleText } from "node:util";
+import { etDate } from "../api/src/srs.ts";
+import { fetchDescription, problemDb, readmeSlugs } from "./db";
+import { ROOT, WINDOWS, bucketDir, reviewQueues, scanBucket, type Bucket } from "./work";
+
+const db = problemDb();
+
+// ── problem index (24h TTL) ──────────────────────────────────────
+
+if (db.indexStale()) {
+  const s = spinner();
+  s.start("Fetching problem index from leetcode.com");
+  try {
+    s.stop(`Indexed ${await db.refreshIndex()} problems`);
+  } catch (err) {
+    if (db.all().length === 0) {
+      s.error(`Fetch failed: ${err}`);
+      process.exit(1);
+    }
+    s.stop(`Index refresh failed (${err}) — using cached copy`);
+  }
+}
+
+// ── description backfill for the campaign problems in README.md ──
+
+const missing = db.missingDescriptions(readmeSlugs(await Bun.file(join(ROOT, "README.md")).text()));
+if (missing.length > 0) {
+  const s = spinner();
+  s.start(`Caching ${missing.length} problem descriptions`);
+  let fetched = 0;
+  let failed = 0;
+  await Promise.all(
+    Array.from({ length: 6 }, async () => {
+      for (let slug = missing.shift(); slug; slug = missing.shift()) {
+        try {
+          const html = await fetchDescription(slug);
+          if (html !== null) db.saveDescription(slug, html);
+          s.message(`Caching problem descriptions (${++fetched})`);
+        } catch {
+          failed++; // left NULL — retried on the next run
+        }
+      }
+    }),
+  );
+  s.stop(failed ? `Cached ${fetched} descriptions (${failed} failed, will retry next run)` : `Cached ${fetched} problem descriptions`);
+}
+
+// ── the three sections ───────────────────────────────────────────
+
+/** What Enter scaffolds: this problem, into this bucket. */
+interface Choice {
+  bucket: Bucket;
+  lc: number;
+  slug: string;
+}
+
+interface Section {
+  label: string;
+  options: { value: Choice; label: string; hint: string }[];
+}
+
+const first = scanBucket(1);
+const problems = db.all();
+const titles = new Map(problems.map((p) => [p.id, p.title] as const));
+db.close();
+
+const queues = reviewQueues(etDate(new Date()));
+
+const sections: Section[] = [
+  {
+    label: "new",
+    options: problems
+      .filter((p) => !first.has(p.id))
+      .map((p) => ({
+        value: { bucket: 1 as Bucket, lc: p.id, slug: p.slug },
+        label: `${p.id}. ${p.title}`,
+        hint: p.paidOnly ? `${p.difficulty} 🔒 premium` : p.difficulty,
+      })),
+  },
+  ...WINDOWS.map((window) => ({
+    label: `${window}-day`,
+    options: queues.get(window)!.map((owed) => ({
+      value: { bucket: window as Bucket, lc: owed.lc, slug: owed.slug },
+      label: `${owed.lc}. ${titles.get(owed.lc) ?? owed.slug}`,
+      hint: `${owed.difficulty} · ${owed.category} · solved ${owed.solved}, ${owed.days}d ago`,
+    })),
+  })),
+];
+
+let active = 0;
+
+/**
+ * The slice of clack's AutocompletePrompt this picker drives. @clack/prompts
+ * re-exports neither the class nor its type, and apps/cli does not depend on
+ * @clack/core directly, so the contract is spelled out structurally.
+ */
+interface Autocomplete {
+  userInput: string;
+  on(event: "key", cb: (char: string | undefined, key: { name?: string }) => void): void;
+  emit(event: "userInput", input: string): void;
+}
+
+/** Search text that cannot match any option — see refresh() below. */
+const NO_MATCH = "\u0000";
+
+let hooked = false;
+
+/**
+ * Bind Tab to the section switch, once per prompt.
+ *
+ * clack has no hook for extra keys, but it calls `options` with the prompt as
+ * `this`, which is all a subscription needs. Switching sections also has to
+ * re-filter the list, and the prompt only recomputes `filteredOptions` when
+ * the search text *changes* — hence the pair of userInput events: an
+ * unmatchable string, then the real one, which recomputes against the new
+ * section, resets the cursor onto its first match and leaves the typed query
+ * on screen untouched. clack repaints as soon as the key handlers return.
+ *
+ * No `placeholder` is passed to the prompt on purpose: clack's own Tab binding
+ * (fill the input with the placeholder) is inert without one.
+ */
+function hookTab(prompt: Autocomplete): void {
+  if (hooked) return;
+  hooked = true;
+  prompt.on("key", (_char, key) => {
+    if (key.name !== "tab") return;
+    active = (active + 1) % sections.length;
+    prompt.emit("userInput", NO_MATCH);
+    prompt.emit("userInput", prompt.userInput);
+  });
+}
+
+// Widened, not asserted: clack resolves `undefined` when Enter lands on an
+// empty list, which its own types do not admit.
+const picked: Choice | symbol | undefined = await autocomplete<Choice>({
+  // Getter, not a string: clack reads `message` on every repaint, so the tab
+  // bar tracks the active section.
+  get message(): string {
+    const bar = sections
+      .map((section, i) => {
+        const text = `${section.label} (${section.options.length})`;
+        return i === active ? styleText("cyan", `▸ ${text}`) : styleText("dim", text);
+      })
+      .join(styleText("dim", "  ·  "));
+    return `Pick a problem   ${bar}   ${styleText("dim", "Tab: next section")}`;
+  },
+  maxItems: 12,
+  options(this: Autocomplete) {
+    hookTab(this);
+    return sections[active]!.options;
+  },
+  filter: (search, option) => {
+    const haystack = option.label!.toLowerCase();
+    return search
+      .toLowerCase()
+      .split(/\s+/)
+      .every((token) => haystack.includes(token));
+  },
+});
+
+if (isCancel(picked)) {
+  cancel("Nothing picked.");
+  process.exit(0);
+}
+if (picked === undefined) {
+  // Enter on an empty section: nothing was focused, so there is nothing to do.
+  cancel(`Nothing owed in the ${sections[active]!.label} section.`);
+  process.exit(0);
+}
+
+// ── scaffold into the picked bucket ──────────────────────────────
+
+// The active workspace names the config file; everything else about it belongs
+// to leetcode-cli.
+const registry = join(homedir(), ".leetcode", "workspaces.json");
+let workspace = "default";
+if (existsSync(registry)) {
+  const parsed: unknown = JSON.parse(readFileSync(registry, "utf8"));
+  if (parsed && typeof parsed === "object" && "active" in parsed && typeof parsed.active === "string") {
+    workspace = parsed.active || workspace;
+  }
+}
+const configPath = join(homedir(), ".leetcode", "workspaces", workspace, "config.json");
+if (!existsSync(configPath)) {
+  log.error(`No leetcode-cli config at ${configPath} — run \`leetcode config -i\` once.`);
+  process.exit(1);
+}
+
+/**
+ * Point leetcode-cli at one bucket. It reads its output directory from the
+ * active workspace config and from nowhere else — no flag, no environment
+ * override — so `workDir` is the whole handover.
+ *
+ * Each write MERGES the file as it is on disk at that moment; a saved copy is
+ * never played back. The CLI owns this file too (theme, language, whatever a
+ * `leetcode login` adds), and replaying stale bytes over it would silently
+ * undo what it wrote while the child ran. After the pick, `workDir` is set to
+ * work/1 instead of to whatever it was before: the first-solve bucket is the
+ * one resting value that is always right, so even a killed run leaves a bare
+ * `leetcode pick` on-layout.
+ */
+function setWorkDir(target: string): void {
+  const raw: unknown = JSON.parse(readFileSync(configPath, "utf8"));
+  if (raw === null || typeof raw !== "object") {
+    log.error(`${configPath} holds no JSON object — run \`leetcode config -i\` to rewrite it.`);
+    process.exit(1);
+  }
+  // 2-space JSON with a trailing newline: leetcode-cli's own format.
+  writeFileSync(configPath, `${JSON.stringify({ ...raw, workDir: target }, null, 2)}\n`);
+}
+
+const dir = bucketDir(picked.bucket);
+mkdirSync(dir, { recursive: true });
+setWorkDir(dir);
+
+let code: number;
+try {
+  const proc = Bun.spawn(
+    // bun installs workspace-dep bins into this package's node_modules, not the root's
+    [join(import.meta.dir, "node_modules", ".bin", "leetcode"), "pick", picked.slug],
+    { cwd: ROOT, stdio: ["inherit", "inherit", "inherit"] },
+  );
+  code = await proc.exited;
+} finally {
+  setWorkDir(bucketDir(1));
+}
+
+// leetcode-cli exits 0 even when it wrote nothing (an expired session only
+// prints a warning), so the scaffold itself is the acceptance test: without a
+// file for this problem in the bucket, the pick failed.
+if (!scanBucket(picked.bucket).has(picked.lc)) {
+  log.error(
+    `leetcode-cli scaffolded nothing for LC ${picked.lc} under work/${picked.bucket}/.\n` +
+      "If it printed “Session expired”, run `leetcode login` and pick again.",
+  );
+  process.exit(code === 0 ? 1 : code);
+}
+process.exit(code);

apps/cli/report.ts

@@ -0,0 +1,37 @@
+/**
+ * Console + GitHub Actions reporting shared by the issue reconcilers.
+ *
+ * Both reconcilers end the same way: an aligned table on stdout, and the same
+ * table as Markdown in the step summary when running under Actions.
+ */
+
+/** Print rows[0] as a header, a rule, then the body — every column padded. */
+export function printTable(rows: string[][]): void {
+  const widths = rows[0]!.map((_, i) => Math.max(...rows.map((r) => r[i]!.length)));
+  const render = (r: string[]) =>
+    r
+      .map((cell, i) => cell.padEnd(widths[i]!))
+      .join("  ")
+      .trimEnd();
+
+  console.log(render(rows[0]!));
+  console.log(widths.map((n) => "─".repeat(n)).join("  "));
+  for (const row of rows.slice(1)) console.log(render(row));
+}
+
+/** Same rows as a GitHub-flavoured Markdown table; rows[0] is the header. */
+export function markdownTable(rows: string[][]): string {
+  const [header, ...body] = rows;
+  return [
+    `| ${header!.join(" | ")} |`,
+    `| ${header!.map(() => "---").join(" | ")} |`,
+    ...body.map((r) => `| ${r.join(" | ")} |`),
+  ].join("\n");
+}
+
+/** Append to the Actions step summary; a no-op outside Actions. */
+export async function writeStepSummary(markdown: string): Promise<void> {
+  const path = process.env.GITHUB_STEP_SUMMARY;
+  if (!path) return;
+  await Bun.write(path, markdown);
+}

apps/cli/solutions.ts

@@ -0,0 +1,115 @@
+/**
+ * Shared work/ solution picker for the test and submit entries: fuzzy-select
+ * a solution file, then hand it to leetcode-cli.
+ *
+ * Every bucket is listed — a 3-day or 7-day re-solve is tested and submitted
+ * exactly like a first solve — and the hint says which one a file belongs to.
+ * The db's per-solution timestamps are keyed by LC number, so the buckets of
+ * one problem share them.
+ *
+ * Every run backfills the db's solutions table with created_at (from the
+ * file's birthtime, mtime when the filesystem has none) so ordering works for
+ * files that predate the db. Ordering is per command — test lists the most
+ * recently tested first (never-tested after, by name); submit lists the most
+ * recently created first. Nothing is ever hidden. Exit 0 of leetcode-cli
+ * stamps last_tested / last_submitted.
+ *
+ * Library: side-effect-free on import — nothing scans the filesystem until
+ * runOnSolution() is called.
+ */
+import { autocomplete, cancel, isCancel, log } from "@clack/prompts";
+import { statSync } from "node:fs";
+import { basename, join, relative } from "node:path";
+import { problemDb } from "./db";
+import { ROOT, WORK as WORK_DIR, parseWorkPath } from "./work";
+
+/** Leading LC number of a work/ file, or undefined for unnumbered names. */
+function lcNumber(rel: string): number | undefined {
+  const m = basename(rel).match(/^(\d+)\./);
+  return m ? Number(m[1]) : undefined;
+}
+
+/** Fuzzy-pick a solution under work/ and run `leetcode <command>` on it. */
+export async function runOnSolution(command: "test" | "submit", message: string): Promise<never> {
+  const db = problemDb();
+
+  const scanned = [...new Bun.Glob("**/*.{js,ts,py,java,c,cpp,go,rs,rb,swift,kt,cs}").scanSync({ cwd: WORK_DIR })];
+
+  // register creation times so files older than the db still sort correctly
+  for (const f of scanned) {
+    const num = lcNumber(f);
+    if (num === undefined) continue;
+    const stat = statSync(join(WORK_DIR, f));
+    db.ensureSolution(num, new Date(stat.birthtimeMs || stat.mtimeMs).toISOString());
+  }
+  const activity = db.solutionActivity();
+
+  // test → most recently tested first; submit → most recently created first
+  const sortKey = (f: string) => {
+    const num = lcNumber(f);
+    const row = num === undefined ? undefined : activity.get(num);
+    return (command === "test" ? row?.lastTested : row?.createdAt) ?? undefined;
+  };
+
+  const files = scanned
+    .sort((a, b) => {
+      const ka = sortKey(a);
+      const kb = sortKey(b);
+      if (ka !== kb) {
+        if (ka === undefined) return 1;
+        if (kb === undefined) return -1;
+        return kb.localeCompare(ka); // ISO timestamps: lexicographic = chronological
+      }
+      return a.localeCompare(b, undefined, { numeric: true });
+    });
+
+  if (files.length === 0) {
+    log.error(`No solution files found in ${relative(process.cwd(), WORK_DIR)}/`);
+    process.exit(1);
+  }
+
+  const picked = await autocomplete<string>({
+    message,
+    placeholder: "Type to search...",
+    maxItems: 12,
+    options: files.map((f) => {
+      const parsed = parseWorkPath(f);
+      const stage = parsed?.bucket === 1 ? "first solve" : `${parsed?.bucket}-day review`;
+      return {
+        value: f,
+        label: basename(f),
+        hint: parsed ? `${stage} · ${parsed.difficulty} · ${parsed.category}` : f,
+      };
+    }),
+    filter: (search, option) => {
+      const haystack = option.value.toLowerCase();
+      return search
+        .toLowerCase()
+        .split(/\s+/)
+        .every((token) => haystack.includes(token));
+    },
+  });
+
+  if (isCancel(picked)) {
+    cancel("Nothing selected.");
+    process.exit(0);
+  }
+
+  const proc = Bun.spawn(
+    [
+      // bun installs workspace-dep bins into this package's node_modules, not the root's
+      join(import.meta.dir, "node_modules", ".bin", "leetcode"),
+      command,
+      join(WORK_DIR, picked),
+    ],
+    { cwd: ROOT, stdio: ["inherit", "inherit", "inherit"] },
+  );
+  const code = await proc.exited;
+  const num = lcNumber(picked);
+  if (code === 0 && num !== undefined) {
+    if (command === "test") db.markTested(num);
+    else db.markSubmitted(num);
+  }
+  db.close();
+  process.exit(code);
+}

apps/cli/source.ts

@@ -0,0 +1,61 @@
+/**
+ * Is a work/ file a real solution or just a scaffolded stub?
+ *
+ * `bun run pick` writes a statement header plus an empty function body, so
+ * file existence alone means nothing: close-solved.ts must not close an issue
+ * on a stub, and work.ts must not ask for the spaced-repetition re-solve of a
+ * problem that was never solved in the first place.
+ *
+ * sync.ts keeps its own copy of the header split — it runs its whole pipeline
+ * on import, so it cannot be imported from. Don't "deduplicate" that one.
+ *
+ * Library: side-effect-free on import.
+ */
+
+/** Placeholder bodies that leetcode-cli / a human leaves behind. */
+const PLACEHOLDERS: Record<string, true> = { pass: true, "...": true, TODO: true };
+
+/**
+ * Decide whether a work/ file contains an implementation.
+ *
+ * The header comment is dropped the same way sync.ts splitSource() does it.
+ * Comments must go before any brace analysis: the scaffold's JSDoc carries
+ * `@param {number[]}`, whose braces would otherwise read as a body.
+ *
+ * A brace-language file counts as implemented when at least one *innermost*
+ * brace pair holds real content. That distinguishes a bare stub
+ * (`function(nums) {}`) and a class-shaped design stub (every method body
+ * empty) from any genuine solution, whose innermost block always has code.
+ * Every non-Python language the campaign has ever used is brace-shaped, so
+ * "not .py" is the only branch that matters.
+ */
+export function isImplemented(src: string, lang: "js" | "py"): boolean {
+  if (lang === "py") {
+    const open = src.indexOf('"""');
+    const close = src.indexOf('"""', open + 3);
+    const code = open === -1 || close === -1 ? src : src.slice(close + 3);
+    for (const raw of code.split("\n")) {
+      const line = raw.replace(/#.*$/, "").trim();
+      if (!line || PLACEHOLDERS[line]) continue;
+      if (/^(?:@|def\s|class\s)/.test(line)) continue;
+      return true; // a statement inside some def body
+    }
+    return false;
+  }
+
+  const headerEnd = src.indexOf("*/");
+  const body = headerEnd === -1 ? src : src.slice(headerEnd + 2);
+  const code = body.replace(/\/\*[\s\S]*?\*\//g, " ").replace(/\/\/.*$/gm, " ");
+
+  let innermost = -1;
+  for (let i = 0; i < code.length; i++) {
+    if (code[i] === "{") {
+      innermost = i;
+    } else if (code[i] === "}" && innermost !== -1) {
+      const inner = code.slice(innermost + 1, i).replace(/[\s;]/g, "");
+      if (inner && !PLACEHOLDERS[inner]) return true;
+      innermost = -1; // measured; the enclosing pair is not innermost
+    }
+  }
+  return false;
+}

apps/cli/spaced-repetition.ts

@@ -0,0 +1,384 @@
+#!/usr/bin/env bun
+/**
+ * Open today's spaced-repetition issue: whatever the SRS ladder has due,
+ * linked straight to LeetCode.
+ *
+ * D1 is the schedule and this script is a renderer. `POST /admin/due` returns
+ * exactly the rows the 8 AM digest mails — reviews standing at the `+3` and
+ * `+7` rungs, oldest first — so the issue and the email can never disagree
+ * about a day, and there is no second scheduler to drift. The rung is also
+ * the `work/<n>` bucket the blind re-solve belongs in, which is the section
+ * `bun run pick` opens on Tab.
+ *
+ * How a problem gets here: closing its issue (a push, a digest tap, a `/done`
+ * comment) enters it on the ladder at `+3` with a concrete next_review, and
+ * passing that review moves it to `+7`. Nothing is "activated" by this script
+ * — it asks what is standing today and writes it down. A day's issue is keyed
+ * by its title (`Spaced Repetition — <Month D, YYYY>`), so re-runs rewrite
+ * that one body instead of stacking duplicates.
+ *
+ * Sunday writes nothing: the ladder never schedules the rest day (WINDOWS of
+ * 3 and 7 exist for exactly that reason, see apps/api/src/srs.ts), so there is
+ * nothing to render. Each rung shows at most WINDOW_CAP problems, with the
+ * overflow listed as optional — the digest's leveller moves that overflow to
+ * later working days the next morning.
+ *
+ * Links: D1 stores titles, not URLs, so each due problem's issue body is read
+ * for the canonical LeetCode link. A body with no URL falls back to a slug
+ * guessed from the title and is flagged in the report, because repo titles are
+ * sometimes abbreviated ("LC 167 · Two Sum II") and a guess can 404.
+ *
+ *   bun apps/cli/spaced-repetition.ts                   # create/update today's issue
+ *   bun apps/cli/spaced-repetition.ts --dry-run         # print the body, touch nothing
+ *   bun apps/cli/spaced-repetition.ts --date=2026-09-02 # pretend it is that ET day
+ *
+ * Auth: GH_TOKEN / GITHUB_TOKEN, else falls back to `gh auth token`.
+ *       SRS_ADMIN_KEY reads the due list; SRS_API overrides the Worker URL for
+ *       local `wrangler dev` runs.
+ */
+import { WINDOWS, etDate, isRestDay } from "../api/src/srs.ts";
+import { github } from "./github.ts";
+import { markdownTable, printTable, writeStepSummary } from "./report.ts";
+
+const DRY =
+  process.argv.includes("--dry-run") ||
+  process.env.DRY_RUN === "1" ||
+  process.env.DRY_RUN === "true";
+
+/**
+ * Required re-solves per window: 3 + 3 on top of the day's 3 new core problems
+ * is already a 90-minute session. A window that closed more than this (the
+ * catalog backfill closed 19 issues in one day) keeps the rest listed under a
+ * collapsed "optional" block rather than pretending they were never due.
+ */
+const WINDOW_CAP = 3;
+
+const LABEL = "spaced-repetition";
+
+// ── ET calendar dates ────────────────────────────────────────────
+
+// Every date here is an ET calendar string (YYYY-MM-DD): the campaign runs on
+// ET days, and an issue closed at 21:30 ET belongs to that ET day, not to the
+// next UTC one. The helpers come from the Worker's SRS domain so this script,
+// the picker and D1 cannot disagree about a date, a window, or the rest day.
+
+const LONG = new Intl.DateTimeFormat("en-US", {
+  timeZone: "UTC",
+  weekday: "long",
+  month: "long",
+  day: "numeric",
+  year: "numeric",
+});
+
+/** `2026-08-26` -> `August 26, 2026`, or with the weekday prefix. */
+function longDate(date: string, weekday = false): string {
+  // Noon-UTC anchor, like every date in srs.ts: no DST edge can move the day.
+  const text = LONG.format(new Date(`${date}T12:00:00Z`));
+  return weekday ? text : text.slice(text.indexOf(", ") + 2);
+}
+
+// ── today ────────────────────────────────────────────────────────
+
+// `||`, not `??`: the workflow passes SRS_TODAY="" on scheduled runs, where the
+// input has no value.
+const override = process.argv.find((a) => a.startsWith("--date="))?.slice(7) || process.env.SRS_TODAY;
+if (override && !/^\d{4}-\d{2}-\d{2}$/.test(override)) {
+  throw new Error(`--date wants YYYY-MM-DD, got ${override}`);
+}
+const today = override || etDate(new Date());
+
+// Sunday is the campaign's rest day and the reason the windows are 3 and 7:
+// nothing is ever booked on it, so there is no issue to write. The cron still
+// fires — the guard lives here, not in the workflow, so a manual run agrees.
+if (isRestDay(today)) {
+  console.log(`${today} is the rest day — no spaced-repetition issue.`);
+  await writeStepSummary(`### Spaced Repetition — ${longDate(today)}\n\nRest day: nothing scheduled.\n`);
+  process.exit(0);
+}
+
+// ── what is due today (D1, via the Worker) ───────────────────────
+
+// The ladder is the schedule: `POST /admin/due` returns exactly the rows the
+// morning digest mails, so the issue and the email can never disagree about a
+// day. This script adds no scheduling of its own — it renders, caps, and
+// links.
+const SRS_API = process.env.SRS_API ?? "https://srs-api.prdlk.workers.dev";
+
+/** One problem the ladder wants re-solved today, as the Worker reports it. */
+interface Due {
+  lc: number;
+  issue: number;
+  title: string;
+  difficulty: string;
+  /** The rung being settled: "+3" or "+7". */
+  stage: string;
+  /** The rung's number — also the `work/<n>` bucket the re-solve belongs in. */
+  bucket: number;
+  due: string;
+  late: number;
+  last_seen: string | null;
+  days_since: number | null;
+}
+
+function readDue(value: unknown): Due | undefined {
+  if (!value || typeof value !== "object") return;
+  const num = (key: string): number | undefined =>
+    key in value && typeof (value as Record<string, unknown>)[key] === "number"
+      ? ((value as Record<string, unknown>)[key] as number)
+      : undefined;
+  const str = (key: string): string | undefined =>
+    key in value && typeof (value as Record<string, unknown>)[key] === "string"
+      ? ((value as Record<string, unknown>)[key] as string)
+      : undefined;
+  const lc = num("lc");
+  const issue = num("issue");
+  const bucket = num("bucket");
+  const stage = str("stage");
+  const title = str("title");
+  const dueOn = str("due");
+  if (lc === undefined || issue === undefined || bucket === undefined) return;
+  if (stage === undefined || title === undefined || dueOn === undefined) return;
+  const difficulty = str("difficulty") ?? "unrated";
+  return {
+    lc,
+    issue,
+    title,
+    difficulty: difficulty.charAt(0).toUpperCase() + difficulty.slice(1),
+    stage,
+    bucket,
+    due: dueOn,
+    late: num("late") ?? 0,
+    last_seen: str("last_seen") ?? null,
+    days_since: num("days_since") ?? null,
+  };
+}
+
+if (!process.env.SRS_ADMIN_KEY) {
+  // Failing loudly beats writing an issue from a second, guessed schedule:
+  // there is only one schedule now, and it lives in D1.
+  console.error("SRS_ADMIN_KEY unset — cannot read today's due list from the Worker.");
+  process.exit(1);
+}
+const res = await fetch(`${SRS_API}/admin/due?date=${today}`, {
+  method: "POST",
+  headers: { authorization: `Bearer ${process.env.SRS_ADMIN_KEY}` },
+});
+if (!res.ok) throw new Error(`GET due list -> ${res.status} ${await res.text()}`);
+const payload: unknown = await res.json();
+const dueRaw: unknown = payload && typeof payload === "object" && "due" in payload ? payload.due : null;
+const due: Due[] = (Array.isArray(dueRaw) ? dueRaw : []).flatMap((r) => {
+  const row = readDue(r);
+  return row ? [row] : [];
+});
+
+// ── repo + auth ──────────────────────────────────────────────────
+
+const gh = await github();
+
+// ── canonical LeetCode links ─────────────────────────────────────
+
+const LC_URL = /https:\/\/leetcode\.com\/problems\/[a-z0-9-]+/;
+
+/** LeetCode's own slug shape, for bodies that lost their URL line. */
+function slugify(name: string): string {
+  return name
+    .toLowerCase()
+    .replace(/[^a-z0-9]+/g, "-")
+    .replace(/^-|-$/g, "");
+}
+
+/**
+ * D1 stores titles, not URLs, so each due problem's issue body is read for the
+ * canonical LeetCode link written by the campaign-issues convention. That is
+ * one request per due problem — at most REVIEW_CAP per rung — instead of the
+ * old walk over every closed issue in the repo. A body with no URL falls back
+ * to a slug guessed from the title and is flagged in the report, because repo
+ * titles are sometimes abbreviated ("LC 167 · Two Sum II") and a guess can 404.
+ */
+interface Linked extends Due {
+  url: string;
+  linked: boolean;
+}
+
+async function link(row: Due): Promise<Linked> {
+  let url: string | undefined;
+  try {
+    const issue: unknown = await gh.api(`/repos/${gh.repo}/issues/${row.issue}`);
+    const body =
+      issue && typeof issue === "object" && "body" in issue && typeof issue.body === "string"
+        ? issue.body
+        : "";
+    url = LC_URL.exec(body)?.[0];
+  } catch (err) {
+    console.warn(`#${row.issue}: could not read the issue body (${err})`);
+  }
+  return {
+    ...row,
+    url: `${url ?? `https://leetcode.com/problems/${slugify(row.title)}`}/`,
+    linked: url !== undefined,
+  };
+}
+
+// ── issue body ───────────────────────────────────────────────────
+
+const title = `Spaced Repetition — ${longDate(today)}`;
+
+/** The window's due problems, capped, oldest scheduled first. */
+const byWindow = new Map<number, Linked[]>();
+for (const days of WINDOWS) {
+  const rows = due.filter((d) => d.bucket === days);
+  byWindow.set(days, await Promise.all(rows.map(link)));
+}
+
+function bucket(days: number): Linked[] {
+  return byWindow.get(days) ?? [];
+}
+
+/** One checkbox line, straight to LeetCode. */
+function line(s: Linked, box: boolean): string {
+  const seen = s.days_since === null ? "unlogged" : `last seen ${s.days_since}d ago`;
+  return (
+    `${box ? "- [ ] " : "- "}[LC ${s.lc} · ${s.title}](${s.url}) · ${s.difficulty} · #${s.issue}` +
+    ` · ${seen}${s.late > 0 ? ` · ⏰ ${s.late}d late` : ""}` +
+    (s.linked ? "" : " · ⚠️ link guessed from title")
+  );
+}
+
+const sections = WINDOWS.map((days) => {
+  const rows = bucket(days);
+  const required = rows.slice(0, WINDOW_CAP);
+  const spill = rows.slice(WINDOW_CAP);
+  return [
+    `### +${days} days — re-solve into \`work/${days}/\``,
+    "",
+    ...(required.length
+      ? required.map((s) => line(s, true))
+      : [`_Nothing at the +${days} rung today._`]),
+    // Overflow is shown, not dropped: the ladder levels everything past the
+    // cap onto later working days, but until that morning's digest runs this
+    // issue is the only record of what was standing.
+    ...(spill.length
+      ? [
+          "",
+          `<details><summary>${spill.length} more at this rung — optional</summary>`,
+          "",
+          ...spill.map((s) => line(s, false)),
+          "",
+          "</details>",
+        ]
+      : []),
+    "",
+  ].join("\n");
+});
+
+const total = WINDOWS.reduce((n, days) => n + Math.min(bucket(days).length, WINDOW_CAP), 0);
+
+const body = [
+  `Re-solve from memory, ${longDate(today, true)}. Blind: no notes, no \`work/\` file, no editor history.`,
+  "",
+  "- 15 minutes per problem, then stop and read your own solution.",
+  "- Say the invariant and the complexity out loud before typing.",
+  "- Failed one? Tick it anyway and expect it back at +3.",
+  "- `bun run pick` → Tab to this rung's section scaffolds the blind re-solve.",
+  "",
+  ...sections,
+  "<sub>Today's due list straight from the SRS ladder in D1 (`POST /admin/due`), rendered by " +
+    "`.github/workflows/spaced-repetition.yml` — re-runs on the same day rewrite this body. " +
+    "The ladder never schedules a Sunday, so there is no issue on the rest day.</sub>",
+].join("\n");
+
+// ── create or refresh today's issue ──────────────────────────────
+
+interface Existing {
+  number: number;
+  title: string;
+  body: string;
+}
+
+function readExisting(value: unknown): Existing | undefined {
+  if (!value || typeof value !== "object") return;
+  if ("pull_request" in value) return;
+  if (!("number" in value) || typeof value.number !== "number") return;
+  if (!("title" in value) || typeof value.title !== "string") return;
+  const text = "body" in value && typeof value.body === "string" ? value.body : "";
+  return { number: value.number, title: value.title, body: text };
+}
+
+// state=all: a day's issue that was already closed must not be recreated.
+let existing: Existing | undefined;
+for await (const raw of gh.list(`/repos/${gh.repo}/issues?labels=${LABEL}&state=all`)) {
+  const row = readExisting(raw);
+  if (row?.title === title) {
+    existing = row;
+    break;
+  }
+}
+
+let status: string;
+if (total === 0) {
+  status = existing
+    ? `nothing due — left #${existing.number} alone`
+    : "nothing due — no issue created";
+} else if (existing && existing.body.trim() === body.trim()) {
+  status = `#${existing.number} already current`;
+} else if (DRY) {
+  status = existing ? `would update #${existing.number}` : "would create";
+} else if (existing) {
+  await gh.api(`/repos/${gh.repo}/issues/${existing.number}`, {
+    method: "PATCH",
+    body: JSON.stringify({ body }),
+  });
+  status = `updated #${existing.number}`;
+} else {
+  // Ensure the label exists before it is used (422 = someone already made it).
+  try {
+    await gh.api(`/repos/${gh.repo}/labels`, {
+      method: "POST",
+      body: JSON.stringify({
+        name: LABEL,
+        color: "0E8A16",
+        description: "Daily blind re-solve of whatever the SRS ladder has due",
+      }),
+    });
+  } catch (err) {
+    if (!String(err).includes("422")) throw err;
+  }
+
+  const created = (await gh.api(`/repos/${gh.repo}/issues`, {
+    method: "POST",
+    body: JSON.stringify({ title, body, labels: [LABEL] }),
+  })) as { number: number };
+  status = `created #${created.number}`;
+}
+
+// ── report ───────────────────────────────────────────────────────
+
+const report = WINDOWS.flatMap((days) =>
+  bucket(days).map((s, i) => [
+    `+${days}d`,
+    String(s.lc),
+    `#${s.issue}`,
+    s.difficulty,
+    s.due + (s.late > 0 ? ` (${s.late}d late)` : ""),
+    i < WINDOW_CAP ? "due" : "optional",
+    s.linked ? s.url : `${s.url} (inferred)`,
+  ]),
+);
+const rows = [["rung", "lc", "issue", "diff", "scheduled", "state", "leetcode"], ...report];
+if (report.length) printTable(rows);
+
+console.log(
+  `\n${title}${DRY ? " (dry run)" : ""}\n` +
+    WINDOWS.map((days) => {
+      const n = bucket(days).length;
+      return `+${days}d: ${Math.min(n, WINDOW_CAP)} due${n > WINDOW_CAP ? ` (+${n - WINDOW_CAP} optional)` : ""}`;
+    }).join(" · ") +
+    `\n${status}`,
+);
+if (DRY) console.log(`\n${body}`);
+
+await writeStepSummary(
+  `### ${title}${DRY ? " (dry run)" : ""}\n\n` +
+    `${total} problem${total === 1 ? "" : "s"} due · ${status}\n\n` +
+    (report.length ? `${markdownTable(rows)}\n` : "The ladder has nothing due today.\n"),
+);

apps/cli/stats.ts

@@ -0,0 +1,458 @@
+#!/usr/bin/env bun
+/**
+ * `bun run stats` — the campaign dashboard, in the terminal.
+ *
+ * Four panes over ONE source: GET /api/stats. D1 owns SRS state, so this
+ * fetches the Worker's aggregate rather than recomputing anything from work/
+ * or from git — a second answer to "how far along am I" is exactly the drift
+ * the campaign is built to avoid. The one local fact shown is the work/ file
+ * inventory, which is inventory, not schedule. Unreachable Worker = an error,
+ * never a reconstructed guess.
+ *
+ * The panes:
+ *   Overview  campaign clock, solved counts, phase bars, recent activity
+ *   Ladder    the +3/+7 rungs and the 14-day due queue
+ *   Concepts  every topic by temperature — what has gone cold
+ *   Activity  the campaign heatmap and the weekly gates
+ *
+ * Outside a TTY (a pipe, CI) it prints every pane once and exits, so the same
+ * command works in a scrollback or a file.
+ */
+import {
+  fetchStats,
+  SRS_API,
+  STAGES,
+  TEMPERATURES,
+  type Stats,
+  type Temperature,
+  type Topic,
+} from "./api.ts";
+import {
+  BLUE,
+  FG,
+  GREEN,
+  GREENS,
+  LINE,
+  MUTED,
+  RED,
+  YELLOW,
+  bar,
+  bold,
+  cell,
+  fg,
+  heatCell,
+  openScreen,
+  padTo,
+  rule,
+  sparkline,
+  stackedBar,
+  width,
+} from "./tui.ts";
+import { BUCKETS, WINDOWS, solvedInBucket } from "./work.ts";
+
+// ── shared vocabulary → colour ───────────────────────────────────
+
+// Temperature is an action ramp, not a thermometer: green is warm enough to
+// leave alone, red is the concept that wants revisiting. The band names carry
+// the heat, the colours carry the urgency — and the glyphs carry both again,
+// so a piped or NO_COLOR dashboard still separates the bands.
+const TEMP_COLOR: Record<Temperature, string> = {
+  hot: GREENS[4]!,
+  fresh: GREEN,
+  fading: YELLOW,
+  cold: RED,
+};
+
+const TEMP_GLYPH: Record<Temperature, string> = {
+  hot: "█",
+  fresh: "▓",
+  fading: "▒",
+  cold: "░",
+};
+
+const DIFF_COLOR: Record<string, string> = { easy: GREEN, medium: YELLOW, hard: RED };
+
+const LABEL = 15; // left-hand label column, shared by every pane
+
+/** "2026-08-31" → "Mon 31", read as a plain calendar date, never localised. */
+function dayLabel(iso: string): string {
+  const [y, m, d] = iso.split("-").map(Number);
+  const weekday = new Date(Date.UTC(y!, m! - 1, d!)).getUTCDay();
+  return `${["Sun", "Mon", "Tue", "Wed", "Thu", "Fri", "Sat"][weekday]} ${String(d).padStart(2, "0")}`;
+}
+
+// ── overview ─────────────────────────────────────────────────────
+
+function overview(s: Stats, cols: number): string[] {
+  const bars = Math.max(10, Math.min(48, cols - LABEL - 18));
+  const lines: string[] = [];
+
+  const { day, days, week } = s.campaign;
+  lines.push(rule("campaign", cols));
+  lines.push(
+    `${cell("Day", LABEL)}${bar(day, days, bars)}  ${fg(FG, `${day}/${days}`)}` +
+      fg(MUTED, `  week ${week}`),
+  );
+  const streak = s.streak > 0 ? fg(GREENS[4]!, `${s.streak} day${s.streak === 1 ? "" : "s"}`) : fg(MUTED, "none");
+  lines.push(`${cell("Streak", LABEL)}${streak}`);
+
+  const total = s.difficulty.reduce((n, d) => n + d.total, 0);
+  const done = s.difficulty.reduce((n, d) => n + d.done, 0);
+  lines.push("");
+  lines.push(rule("solved", cols));
+  lines.push(
+    `${cell("All", LABEL)}${bar(done, total, bars)}  ${fg(FG, `${done}/${total}`)}` +
+      fg(MUTED, `  ${total > 0 ? Math.round((done / total) * 100) : 0}%`),
+  );
+  for (const d of s.difficulty) {
+    lines.push(
+      `${cell(d.difficulty, LABEL)}${bar(d.done, d.total, bars, DIFF_COLOR[d.difficulty] ?? GREEN)}` +
+        `  ${fg(MUTED, `${d.done}/${d.total}`)}`,
+    );
+  }
+
+  lines.push("");
+  lines.push(rule("phases", cols));
+  lines.push(
+    `${" ".repeat(LABEL)}${fg(GREEN, "█")} ${fg(MUTED, "core")}  ${fg(BLUE, "█")} ${fg(MUTED, "optional")}  ` +
+      `${fg(RED, "█")} ${fg(MUTED, "deferred")}  ${fg(LINE, "░")} ${fg(MUTED, "remaining")}`,
+  );
+  // Every phase bar is scaled to the LARGEST phase, not to itself: phases are
+  // not the same size, and equal-length bars would hide that.
+  const scale = Math.max(
+    1,
+    ...s.phases.map((p) => p.core_total + p.optional_total + p.deferred_total),
+  );
+  for (const p of s.phases) {
+    const phaseDone = p.core_done + p.optional_done + p.deferred_done;
+    const phaseTotal = p.core_total + p.optional_total + p.deferred_total;
+    const segments = [
+      { value: p.core_done, color: GREEN },
+      { value: p.optional_done, color: BLUE },
+      { value: p.deferred_done, color: RED },
+      { value: phaseTotal - phaseDone, color: LINE, glyph: "░" },
+    ];
+    lines.push(
+      `${cell(p.name, LABEL - 1)} ${stackedBar(segments, scale, bars)}  ${fg(MUTED, `${phaseDone}/${phaseTotal}`)}`,
+    );
+  }
+
+  lines.push("");
+  lines.push(rule("last 14 days", cols));
+  const window = s.heat.slice(Math.max(0, day - 14), day);
+  const attempts = window.reduce((n, h) => n + h.attempts, 0);
+  const passes = window.reduce((n, h) => n + h.passes, 0);
+  lines.push(
+    `${cell("Attempts", LABEL)}${sparkline(window.map((h) => h.attempts))}  ` +
+      fg(MUTED, `${attempts} attempts · ${passes} passed`),
+  );
+
+  // The one local fact on the dashboard: solutions on disk, per
+  // spaced-repetition bucket. Inventory, not schedule — D1's ladder is the
+  // schedule, and the Ladder pane is where it is shown. Empty scaffolds do not
+  // count, the same rule close-solved.ts closes issues by.
+  lines.push("");
+  lines.push(rule("work/ on disk", cols));
+  lines.push(
+    cell("Solutions", LABEL) +
+      BUCKETS.map((b) => `${fg(FG, String(solvedInBucket(b)))} ${fg(MUTED, `in work/${b}`)}`).join(
+        fg(MUTED, " · "),
+      ),
+  );
+
+  return lines;
+}
+
+// ── ladder ───────────────────────────────────────────────────────
+
+function ladder(s: Stats, cols: number): string[] {
+  const bars = Math.max(10, Math.min(48, cols - LABEL - 18));
+  const lines: string[] = [rule("ladder", cols)];
+  lines.push(fg(MUTED, `${" ".repeat(LABEL)}the review each problem owes next`));
+
+  const rungs = STAGES.map((stage) => s.ladder[stage] ?? 0);
+  const peak = Math.max(1, ...rungs);
+  STAGES.forEach((stage, i) => {
+    const n = rungs[i]!;
+    // new is the untouched pile and retired is the finished one; only the
+    // middle rungs are live work, so only they get the live colour.
+    const color = stage === "new" ? LINE : stage === "retired" ? GREENS[4]! : GREEN;
+    lines.push(`${cell(stage, LABEL)}${bar(n, peak, bars, color)}  ${fg(FG, String(n))}`);
+  });
+
+  lines.push("");
+  lines.push(rule("due queue", cols));
+  const today = s.queue[0];
+  lines.push(
+    fg(MUTED, `${" ".repeat(LABEL)}cumulative: everything the ladder wants by that day`) +
+      (today ? fg(FG, `  ·  ${today.due} due today`) : ""),
+  );
+  const depth = Math.max(1, ...s.queue.map((q) => q.due));
+  for (const q of s.queue) {
+    const color = q.due > 3 ? RED : q.due > 0 ? GREEN : LINE;
+    lines.push(
+      `${cell(dayLabel(q.date), LABEL)}${bar(q.due, depth, bars, color)}  ${fg(MUTED, String(q.due))}`,
+    );
+  }
+
+  return lines;
+}
+
+// ── concepts ─────────────────────────────────────────────────────
+
+/** Coldest first, and everything the schedule has not reached yet last. */
+function byUrgency(a: Topic, b: Topic): number {
+  if (a.taught !== b.taught) return a.taught ? -1 : 1;
+  return (
+    TEMPERATURES.indexOf(b.temperature) - TEMPERATURES.indexOf(a.temperature) ||
+    (b.age ?? Infinity) - (a.age ?? Infinity) ||
+    a.issue - b.issue
+  );
+}
+
+// Row layout: "● " + name + temp + coverage bar + done/total + last + due.
+// Everything but the name is fixed, so the name absorbs the terminal width.
+const TEMP_W = 7;
+const COVER_W = 10;
+const COUNT_W = 6;
+const AGE_W = 6;
+const DUE_W = 3;
+const CONCEPT_FIXED = 3 + (TEMP_W + 1) + (COVER_W + 1) + (COUNT_W + 1) + (AGE_W + 1) + DUE_W + 1;
+
+function concepts(s: Stats, cols: number): string[] {
+  const lines: string[] = [rule("concepts by temperature", cols)];
+  // Thresholds are the ladder's windows; print them rather than restate them.
+  const bands: [Temperature, string][] = [
+    ["hot", `≤${WINDOWS[0]}d`],
+    ["fresh", `≤${WINDOWS[1]}d`],
+    ["fading", `≤${WINDOWS[1] * 2}d`],
+    ["cold", "older, or never"],
+  ];
+  lines.push(
+    fg(MUTED, "  last exercised — ") +
+      bands
+        .map(([t, when]) => `${fg(TEMP_COLOR[t], `${TEMP_GLYPH[t]} ${t}`)} ${fg(MUTED, when)}`)
+        .join(fg(MUTED, " · ")),
+  );
+
+  const taught = s.topics.filter((t) => t.taught);
+  const counts = TEMPERATURES.map((t) => ({
+    value: taught.filter((x) => x.temperature === t).length,
+    color: TEMP_COLOR[t],
+    glyph: TEMP_GLYPH[t],
+  }));
+  lines.push("");
+  lines.push(`  ${stackedBar(counts, Math.max(1, taught.length), Math.max(10, Math.min(60, cols - 4)))}`);
+  lines.push(
+    `  ${TEMPERATURES.map((t, i) => `${fg(TEMP_COLOR[t], TEMP_GLYPH[t])} ${counts[i]!.value} ${fg(MUTED, t)}`).join("  ")}` +
+      fg(MUTED, `   of ${taught.length} taught so far`),
+  );
+
+  const nameW = Math.max(12, Math.min(34, cols - CONCEPT_FIXED));
+  lines.push("");
+  lines.push(
+    fg(
+      MUTED,
+      `   ${cell("concept", nameW)} ${cell("temp", TEMP_W)} ` +
+        `${cell("coverage", COVER_W + 1 + COUNT_W)} ${cell("last", AGE_W)} ${cell("due", DUE_W, "right")}`,
+    ),
+  );
+
+  for (const t of s.topics.slice().sort(byUrgency)) {
+    // An unreached concept has no temperature to report — it shows the week
+    // the schedule gets to it instead, so the list stays one list.
+    const color = t.taught ? TEMP_COLOR[t.temperature] : LINE;
+    const age =
+      !t.taught ? "—" : t.age === null ? "never" : t.age <= 0 ? "today" : `${t.age}d`;
+    lines.push(
+      ` ${fg(color, t.taught ? TEMP_GLYPH[t.temperature] : "·")} ${fg(t.taught ? FG : MUTED, cell(t.name, nameW))} ` +
+        `${fg(color, cell(t.taught ? t.temperature : `wk ${t.week ?? "?"}`, TEMP_W))} ` +
+        `${bar(t.done, t.total, COVER_W, t.taught ? GREEN : LINE)} ` +
+        `${fg(MUTED, cell(`${t.done}/${t.total}`, COUNT_W))} ${fg(MUTED, cell(age, AGE_W))} ` +
+        `${t.due > 0 ? fg(GREEN, cell(String(t.due), DUE_W, "right")) : fg(LINE, cell("·", DUE_W, "right"))}`,
+    );
+  }
+
+  return lines;
+}
+
+// ── activity ─────────────────────────────────────────────────────
+
+const WEEKDAYS = ["Mon", "Tue", "Wed", "Thu", "Fri", "Sat", "Sun"];
+
+function activity(s: Stats, cols: number): string[] {
+  const lines: string[] = [rule("attempts per campaign day", cols)];
+  lines.push(fg(MUTED, "  the README heatmap, same ramp — Sunday is the rest day"));
+  lines.push("");
+
+  // CAMPAIGN_START is a Monday, so the grid is simply the series in weeks of
+  // seven: column = week, row = weekday, exactly as the SVG lays it out.
+  const weeks = Math.ceil(s.heat.length / 7);
+  lines.push(
+    `      ${fg(MUTED, Array.from({ length: weeks }, (_, w) => ` ${String(w + 1).padStart(2)}`).join(""))}`,
+  );
+  WEEKDAYS.forEach((name, row) => {
+    const cells = Array.from({ length: weeks }, (_, w) => {
+      const entry = s.heat[w * 7 + row];
+      if (entry === undefined) return "   ";
+      // The marker sits in the gutter, not inside the cell: a foreground
+      // escape within a background run would end the background early.
+      const marker = s.campaign.day - 1 === w * 7 + row ? fg(FG, "▸") : " ";
+      return marker + heatCell(entry.attempts, "  ");
+    }).join("");
+    lines.push(`  ${fg(MUTED, cell(name, 4))}${cells}`);
+  });
+  lines.push("");
+  lines.push(
+    `  ${fg(MUTED, "less ")}${GREENS.map((_, level) => heatCell(level, "  ")).join("")}${fg(MUTED, " more")}` +
+      fg(MUTED, "   · one cell per day, colour = attempts, ▸ marks today"),
+  );
+
+  lines.push("");
+  lines.push(rule("weekly gates", cols));
+  if (s.gates.length === 0) {
+    lines.push(fg(MUTED, "  no gate has been scored yet"));
+  }
+  const gateW = Math.max(10, Math.min(24, cols - 30));
+  for (const g of s.gates) {
+    // Gates were stored as an integer percent before they were stored as a
+    // fraction; both are in the table, so normalise on read.
+    const rate = g.pass_rate === null ? null : g.pass_rate <= 1 ? g.pass_rate : g.pass_rate / 100;
+    const color = rate === null ? LINE : rate >= 0.7 ? GREEN : RED;
+    lines.push(
+      `  ${fg(FG, cell(`Week ${g.week}`, 9))}` +
+        `${bar(rate ?? 0, 1, gateW, color)}` +
+        `  ${fg(color, cell(rate === null ? "open" : `${Math.round(rate * 100)}%`, 5))}` +
+        `  ${fg(MUTED, g.issue === null ? "" : `#${g.issue}`)}${fg(MUTED, g.closed_on ? `  closed ${g.closed_on}` : "")}`,
+    );
+  }
+
+  return lines;
+}
+
+// ── panes ────────────────────────────────────────────────────────
+
+const PANES: { name: string; render: (s: Stats, cols: number) => string[] }[] = [
+  { name: "Overview", render: overview },
+  { name: "Ladder", render: ladder },
+  { name: "Concepts", render: concepts },
+  { name: "Activity", render: activity },
+];
+
+// ── non-interactive: every pane, once ────────────────────────────
+
+let stats: Stats;
+try {
+  stats = await fetchStats();
+} catch (err) {
+  console.error(`stats: ${err instanceof Error ? err.message : String(err)}`);
+  console.error(`stats: D1 is the only source of SRS truth — nothing to fall back to (SRS_API=${SRS_API}).`);
+  process.exit(1);
+}
+
+if (!process.stdout.isTTY) {
+  const cols = 80;
+  for (const pane of PANES) {
+    console.log(bold(pane.name.toUpperCase()));
+    for (const line of pane.render(stats, cols)) console.log(line);
+    console.log("");
+  }
+  process.exit(0);
+}
+
+// ── interactive ──────────────────────────────────────────────────
+
+let pane = 0;
+let offset = 0;
+let note = "";
+
+// Quitting resolves this; the finally below is what restores the terminal, so
+// a thrown render never leaves the shell in raw mode on the alternate screen.
+const { promise: quit, resolve: leave } = Promise.withResolvers<void>();
+const screen = openScreen(onKey, draw);
+
+function frame(): string[] {
+  const cols = screen.columns;
+  const tabs = PANES.map((p, i) =>
+    i === pane ? bold(fg(FG, ` ${p.name} `)) : fg(MUTED, ` ${p.name} `),
+  ).join(fg(LINE, "│"));
+  const clock = fg(MUTED, `day ${stats.campaign.day}/${stats.campaign.days} · streak ${stats.streak}  `);
+  const header = padTo(tabs, Math.max(0, cols - width(clock))) + clock;
+
+  // header + rule + body + footer fills the terminal exactly.
+  const body = PANES[pane]!.render(stats, cols);
+  const height = Math.max(1, screen.rows - 3);
+  offset = Math.max(0, Math.min(offset, Math.max(0, body.length - height)));
+  const scrolled =
+    body.length > height ? ` · ${offset + 1}-${Math.min(offset + height, body.length)}/${body.length}` : "";
+
+  return [
+    header,
+    fg(LINE, "─".repeat(cols)),
+    ...body.slice(offset, offset + height),
+    fg(MUTED, `Tab/←→ pane · ↑↓ scroll · r refresh · q quit${scrolled}`) + (note ? `  ${fg(YELLOW, note)}` : ""),
+  ];
+}
+
+function draw(): void {
+  screen.draw(frame());
+}
+
+function onKey(chunk: string): void {
+  switch (chunk) {
+    // Raw mode delivers Ctrl-C as a byte, not a signal, so it is handled here.
+    case "q":
+    case "\x03":
+    case "\x1b":
+      leave();
+      return;
+    case "\t":
+    case "\x1b[C":
+    case "l":
+      pane = (pane + 1) % PANES.length;
+      offset = 0;
+      break;
+    case "\x1b[Z":
+    case "\x1b[D":
+    case "h":
+      pane = (pane + PANES.length - 1) % PANES.length;
+      offset = 0;
+      break;
+    case "\x1b[A":
+    case "k":
+      offset = Math.max(0, offset - 1);
+      break;
+    case "\x1b[B":
+    case "j":
+      offset += 1;
+      break;
+    case "r":
+      note = "refreshing…";
+      draw();
+      void fetchStats()
+        .then((fresh) => {
+          stats = fresh;
+          note = "";
+        })
+        .catch((err: unknown) => {
+          note = `refresh failed: ${err instanceof Error ? err.message : String(err)}`;
+        })
+        .finally(draw);
+      return;
+    default:
+      if (chunk >= "1" && chunk <= String(PANES.length)) {
+        pane = Number(chunk) - 1;
+        offset = 0;
+        break;
+      }
+      return;
+  }
+  draw();
+}
+
+try {
+  draw();
+  await quit;
+} finally {
+  screen.close();
+}

apps/cli/submit.ts

@@ -0,0 +1,7 @@
+#!/usr/bin/env bun
+/**
+ * Fuzzy-select a solution file from work/ and submit it to LeetCode's judge.
+ */
+import { runOnSolution } from "./solutions";
+
+await runOnSolution("submit", "Submit which solution?");

apps/cli/sync-project-fields.ts

@@ -0,0 +1,252 @@
+#!/usr/bin/env bun
+/**
+ * Reconcile the "Interview Prep" Project's Set and Difficulty fields from the
+ * labels on each problem issue.
+ *
+ * The issue labels are the truth: `set:core|optional|deferred` and
+ * `diff:easy|medium|hard` are fixed when a problem issue is filed, and the
+ * Project's single-selects are a projection of them so the board can group and
+ * filter. This recomputes every desired value from scratch and writes only the
+ * ones that differ, so re-runs are no-ops and a backfill needs no special
+ * casing — same reconcile-don't-react shape as close-solved / close-topics.
+ *
+ * Scope guard: only items whose issue carries the `problem` label are touched.
+ * Topic, cadence and review issues have no set or difficulty, so they are
+ * skipped outright — never written, never cleared.
+ *
+ * Deliberately NOT mirrored: SRS stage and first-attempt result. Those live
+ * only in D1 (apps/api owns them; the charts, digest and /api/stats read them
+ * from there). The Project mirror writes Target Date and nothing else.
+ *
+ * Auth: needs a token carrying the `project` scope — neither the default
+ * `gh auth token` nor Actions' GITHUB_TOKEN has it. Pass one in GH_TOKEN: the
+ * PROJECT_PAT secret in CI, or the Worker's GH_PAT locally:
+ *   GH_TOKEN=$GH_PAT bun run sync-project-fields -- --dry-run
+ */
+import { github, type GitHub } from "./github.ts";
+import { markdownTable, printTable, writeStepSummary } from "./report.ts";
+
+const PROJECT_TITLE = "Interview Prep";
+
+/** Label → single-select option name, per field. Static tables, so Record. */
+const FIELDS: Record<string, Record<string, string>> = {
+  Set: { "set:core": "Core", "set:optional": "Optional", "set:deferred": "Deferred" },
+  Difficulty: { "diff:easy": "Easy", "diff:medium": "Medium", "diff:hard": "Hard" },
+};
+
+const DRY =
+  process.argv.includes("--dry-run") ||
+  process.env.DRY_RUN === "1" ||
+  process.env.DRY_RUN === "true";
+
+// ── project + field resolution ───────────────────────────────────
+
+interface SelectField {
+  id: string;
+  /** option name → option id */
+  options: Record<string, string>;
+}
+
+interface Project {
+  id: string;
+  fields: Record<string, SelectField>;
+}
+
+/**
+ * Field and option IDs are resolved by NAME, so re-creating a field in the UI
+ * does not strand this script on a dead id.
+ */
+async function readProject(gh: GitHub, owner: string): Promise<Project> {
+  const data = (await gh.graphql(
+    `query($owner: String!, $title: String!) {
+      user(login: $owner) {
+        projectsV2(first: 10, query: $title) {
+          nodes {
+            id title
+            fields(first: 30) {
+              nodes {
+                ... on ProjectV2SingleSelectField { id name options { id name } }
+              }
+            }
+          }
+        }
+      }
+    }`,
+    { owner, title: PROJECT_TITLE },
+  )) as {
+    user: {
+      projectsV2: {
+        nodes: {
+          id: string;
+          title: string;
+          fields: { nodes: ({ id: string; name: string; options: { id: string; name: string }[] } | Record<string, never>)[] };
+        }[];
+      };
+    };
+  };
+
+  const node = data.user.projectsV2.nodes.find((n) => n.title === PROJECT_TITLE);
+  if (!node) throw new Error(`project "${PROJECT_TITLE}" not found for @${owner}`);
+
+  const fields: Record<string, SelectField> = {};
+  for (const name of Object.keys(FIELDS)) {
+    const field = node.fields.nodes.find((f) => "name" in f && f.name === name);
+    if (!field || !("options" in field)) {
+      throw new Error(`single-select field "${name}" missing from "${PROJECT_TITLE}"`);
+    }
+    fields[name] = {
+      id: field.id,
+      options: Object.fromEntries(field.options.map((o) => [o.name, o.id])),
+    };
+  }
+  return { id: node.id, fields };
+}
+
+// ── items ────────────────────────────────────────────────────────
+
+interface Item {
+  id: string;
+  issue: number;
+  title: string;
+  labels: string[];
+  /** field name → current option name */
+  current: Record<string, string>;
+}
+
+async function* readItems(gh: GitHub, project: string): AsyncGenerator<Item, void, void> {
+  let cursor: string | null = null;
+  for (;;) {
+    const data = (await gh.graphql(
+      `query($id: ID!, $after: String) {
+        node(id: $id) {
+          ... on ProjectV2 {
+            items(first: 100, after: $after) {
+              pageInfo { hasNextPage endCursor }
+              nodes {
+                id
+                fieldValues(first: 25) {
+                  nodes {
+                    ... on ProjectV2ItemFieldSingleSelectValue {
+                      name field { ... on ProjectV2FieldCommon { name } }
+                    }
+                  }
+                }
+                content {
+                  ... on Issue { number title labels(first: 25) { nodes { name } } }
+                }
+              }
+            }
+          }
+        }
+      }`,
+      { id: project, after: cursor },
+    )) as {
+      node: {
+        items: {
+          pageInfo: { hasNextPage: boolean; endCursor: string };
+          nodes: {
+            id: string;
+            fieldValues: { nodes: ({ name: string; field: { name: string } } | Record<string, never>)[] };
+            content: { number: number; title: string; labels: { nodes: { name: string }[] } } | Record<string, never>;
+          }[];
+        };
+      };
+    };
+
+    for (const node of data.node.items.nodes) {
+      const content = node.content;
+      // Draft items and PRs have no issue number; nothing to reconcile.
+      if (!("number" in content) || typeof content.number !== "number") continue;
+      const current: Record<string, string> = {};
+      for (const value of node.fieldValues.nodes) {
+        if ("name" in value && value.field?.name) current[value.field.name] = value.name;
+      }
+      yield {
+        id: node.id,
+        issue: content.number,
+        title: content.title,
+        labels: content.labels.nodes.map((l) => l.name),
+        current,
+      };
+    }
+
+    const page = data.node.items.pageInfo;
+    if (!page.hasNextPage) return;
+    cursor = page.endCursor;
+  }
+}
+
+// ── reconcile ────────────────────────────────────────────────────
+
+interface Write {
+  item: string;
+  issue: number;
+  title: string;
+  field: string;
+  from: string;
+  to: string;
+}
+
+async function setField(gh: GitHub, project: string, w: Write, option: string): Promise<void> {
+  await gh.graphql(
+    `mutation($project: ID!, $item: ID!, $field: ID!, $option: String!) {
+      updateProjectV2ItemFieldValue(
+        input: {
+          projectId: $project, itemId: $item, fieldId: $field
+          value: { singleSelectOptionId: $option }
+        }
+      ) { projectV2Item { id } }
+    }`,
+    { project, item: w.item, field: w.field, option },
+  );
+}
+
+const gh = await github();
+const owner = gh.repo.split("/")[0]!;
+const project = await readProject(gh, owner);
+
+const writes: Write[] = [];
+let problems = 0;
+let skipped = 0;
+
+for await (const item of readItems(gh, project.id)) {
+  if (!item.labels.includes("problem")) {
+    skipped++;
+    continue;
+  }
+  problems++;
+  for (const [field, mapping] of Object.entries(FIELDS)) {
+    const want = item.labels.map((l) => mapping[l]).find(Boolean);
+    if (!want) {
+      console.warn(`#${item.issue} has no ${field.toLowerCase()} label — left untouched`);
+      continue;
+    }
+    const from = item.current[field];
+    if (from === want) continue;
+    if (!project.fields[field]!.options[want]) {
+      throw new Error(`field "${field}" has no option "${want}"`);
+    }
+    writes.push({
+      item: item.id,
+      issue: item.issue,
+      title: item.title,
+      field: project.fields[field]!.id,
+      from: from ?? "(empty)",
+      to: want,
+    });
+  }
+}
+
+const rows: string[][] = [["ISSUE", "PROBLEM", "FIELD", "FROM", "TO"]];
+for (const w of writes) {
+  const name = Object.entries(project.fields).find(([, f]) => f.id === w.field)![0];
+  rows.push([`#${w.issue}`, w.title.slice(0, 46), name, w.from, w.to]);
+  if (DRY) continue;
+  await setField(gh, project.id, w, project.fields[name]!.options[w.to]!);
+}
+
+if (writes.length) printTable(rows);
+console.log(
+  `\n${DRY ? "would update" : "updated"} ${writes.length} field value(s) across ${problems} problem item(s); ${skipped} non-problem item(s) skipped`,
+);
+if (writes.length) await writeStepSummary(markdownTable(rows));

apps/cli/sync.ts

@@ -1,6 +1,6 @@
 #!/usr/bin/env bun
 /**
- * Sync work/ leetcode solutions into docs/solutions/ pages.
+ * Sync work/1 leetcode solutions into apps/docs/content/ pages.
  *
  * - New problems get a full page (frontmatter, badge, warning, examples,
  *   constraints, solution) in the gold-standard format.
@@ -9,13 +9,17 @@
  * - When a problem is solved in both JavaScript and Python the solution
  *   renders as a <CodeGroup> (Python first); single-language solutions
  *   render as a plain fence.
+ *
+ * Only the first-solve bucket is read: work/3 and work/7 hold blind re-solves
+ * of problems whose page already exists (see work.ts), and a fresh scaffold
+ * there would otherwise wipe the page's `## Solution`.
  */
 import { mkdir, rename } from "node:fs/promises";
 import { basename, join } from "node:path";
 
-const ROOT = join(import.meta.dir, "..");
-const WORK = join(ROOT, "work");
-const DOCS = join(ROOT, "docs", "solutions");
+const ROOT = join(import.meta.dir, "..", "..");
+const WORK = join(ROOT, "work", "1");
+const DOCS = join(ROOT, "apps", "docs", "content");
 
 type Lang = "js" | "py";
 
@@ -323,7 +327,7 @@ for await (const rel of new Bun.Glob("**/*.{js,py}").scan(WORK)) {
   const file = basename(rel);
   const m = file.match(/^(\d+)\.(.+)\.(?:js|py)$/);
   if (!m) {
-    console.warn(`skip (unrecognized name): work/${rel}`);
+    console.warn(`skip (unrecognized name): work/1/${rel}`);
     continue;
   }
   const num = Number(m[1]);
@@ -355,6 +359,24 @@ function ensureLabel(page: string, h: Header): string {
   return page.replace(/^sidebar:$/m, `sidebar:\n${label}`);
 }
 
+/**
+ * Offset just past the script-owned `## Solution` section: the next top-level
+ * heading that is not inside a code fence, or end of file. Sections a human
+ * appends after the solution (`## Explanation`, …) survive regeneration.
+ */
+function solutionEnd(page: string, from: number): number {
+  const lines = page.slice(from).split("\n");
+  let offset = from + lines[0]!.length + 1; // past the '## Solution' line itself
+  let fenced = false;
+  for (let i = 1; i < lines.length; i++) {
+    const line = lines[i]!;
+    if (line.startsWith("```")) fenced = !fenced;
+    else if (!fenced && line.startsWith("## ")) return offset;
+    offset += line.length + 1;
+  }
+  return page.length;
+}
+
 // Existing docs pages, keyed by leetcode number from the frontmatter title.
 const pages = new Map<number, string>(); // num -> path relative to DOCS
 for (const rel of (await Array.fromAsync(new Bun.Glob("**/*.mdx").scan(DOCS))).sort()) {
@@ -382,7 +404,9 @@ for (const sol of ordered) {
       report.push([workRef, existing, "ERROR: no '## Solution' heading"]);
       continue;
     }
-    const next = ensureLabel(page.slice(0, at) + solutionSection(sol), sol.header);
+    const tail = page.slice(solutionEnd(page, at));
+    const body = page.slice(0, at) + solutionSection(sol) + (tail ? `\n${tail}` : "");
+    const next = ensureLabel(body, sol.header);
     const actions: string[] = [];
     if (existing !== name) {
       await rename(join(DOCS, existing), join(DOCS, name));

apps/cli/test.ts

@@ -0,0 +1,7 @@
+#!/usr/bin/env bun
+/**
+ * Fuzzy-select a solution file from work/ and run leetcode-cli tests on it.
+ */
+import { runOnSolution } from "./solutions";
+
+await runOnSolution("test", "Test which solution?");

apps/cli/tui.ts

@@ -0,0 +1,214 @@
+/**
+ * Terminal drawing primitives for the stats dashboard: colour, block-glyph
+ * charts, and the alternate-screen plumbing.
+ *
+ * Hand-rolled for the same reason apps/api/src/charts.ts and png.ts are — the
+ * repo carries near-zero runtime dependencies, and everything a progress
+ * dashboard needs is a bar, a sparkline and a coloured cell. The palette is
+ * charts.ts's, so the terminal, the README SVGs and the digest email are the
+ * same picture in three renderers.
+ *
+ * Side-effect-free on import except for reading the colour environment once;
+ * nothing is written to the terminal until openScreen() is called.
+ */
+
+// ── palette (apps/api/src/charts.ts) ─────────────────────────────
+
+export const FG = "#fafafa"; // zinc-50 text
+export const MUTED = "#a1a1aa"; // zinc-400 secondary text
+export const LINE = "#27272a"; // zinc-800 structure / empty track
+export const GREEN = "#16a34a"; // green-600 (core / pass)
+export const BLUE = "#2563eb"; // blue-600 (optional)
+export const YELLOW = "#eab308"; // yellow-500 (attention)
+export const RED = "#f87171"; // red-400 (fail, on dark)
+
+/** GitHub dark-mode contribution hues; level 0 is the empty cell. */
+export const GREENS = ["#27272a", "#0e4429", "#006d32", "#26a641", "#39d353"];
+
+// ── colour ───────────────────────────────────────────────────────
+
+/**
+ * Honour NO_COLOR and pipes, and let FORCE_COLOR override both — the dashboard
+ * is legible without colour, and `bun run stats | less -R` should still be.
+ */
+export const COLOR =
+  process.env.FORCE_COLOR !== undefined
+    ? process.env.FORCE_COLOR !== "0"
+    : process.env.NO_COLOR === undefined && process.stdout.isTTY === true;
+
+const RESET = "\x1b[0m";
+
+function channels(hex: string): string {
+  return `${parseInt(hex.slice(1, 3), 16)};${parseInt(hex.slice(3, 5), 16)};${parseInt(hex.slice(5, 7), 16)}`;
+}
+
+export function fg(hex: string, text: string): string {
+  return COLOR ? `\x1b[38;2;${channels(hex)}m${text}${RESET}` : text;
+}
+
+export function bg(hex: string, text: string): string {
+  return COLOR ? `\x1b[48;2;${channels(hex)}m${text}${RESET}` : text;
+}
+
+export function bold(text: string): string {
+  return COLOR ? `\x1b[1m${text}${RESET}` : text;
+}
+
+const ANSI = /\x1b\[[0-9;]*m/g;
+
+/** Printable width — every glyph used here is one cell wide. */
+export function width(text: string): number {
+  return text.replace(ANSI, "").length;
+}
+
+/** Pad or cut a *plain* string to exactly n cells; colour it afterwards. */
+export function cell(text: string, n: number, align: "left" | "right" = "left"): string {
+  const clipped = text.length > n ? `${text.slice(0, Math.max(0, n - 1))}…` : text;
+  return align === "left" ? clipped.padEnd(n) : clipped.padStart(n);
+}
+
+/** Pad an already-coloured string to n cells. */
+export function padTo(text: string, n: number): string {
+  return text + " ".repeat(Math.max(0, n - width(text)));
+}
+
+// ── charts ───────────────────────────────────────────────────────
+
+const EIGHTHS = ["", "▏", "▎", "▍", "▌", "▋", "▊", "▉"];
+const SPARKS = ["▁", "▂", "▃", "▄", "▅", "▆", "▇", "█"];
+
+/**
+ * Filled and empty. The two differ by GLYPH as well as by colour, so a bar
+ * still reads under NO_COLOR or in a pipe — every chart here has to survive
+ * `bun run stats > notes.txt`.
+ */
+const FILL = "█";
+const TRACK = "░";
+
+/**
+ * One bar over a muted track, `cells` wide. Sub-cell remainders use the
+ * eighth-block glyphs, so one solved problem out of 161 still shows.
+ */
+export function bar(value: number, total: number, cells: number, color = GREEN): string {
+  const units = Math.round(Math.max(0, Math.min(1, total > 0 ? value / total : 0)) * cells * 8);
+  const filled = FILL.repeat(Math.floor(units / 8)) + EIGHTHS[units % 8]!;
+  return fg(color, filled) + fg(LINE, TRACK.repeat(Math.max(0, cells - filled.length)));
+}
+
+export interface Segment {
+  value: number;
+  color: string;
+  /** Defaults to the solid block; pass TRACK-ish glyphs for "not done yet". */
+  glyph?: string;
+}
+
+/**
+ * Segments laid end to end, `total` mapped to `cells`. Anything short of
+ * `total` is left BLANK, not tracked: bars sharing a scale must show their
+ * different lengths. A caller wanting a visible remainder passes it as a
+ * segment of its own.
+ *
+ * Non-zero segments are floored but never vanish; the overshoot that floor
+ * creates is taken back from the widest segment, which can spare it.
+ */
+export function stackedBar(segments: Segment[], total: number, cells: number): string {
+  if (total <= 0) return " ".repeat(cells);
+  const widths = segments.map((s) =>
+    s.value > 0 ? Math.max(1, Math.floor((s.value / total) * cells)) : 0,
+  );
+  let over = widths.reduce((a, b) => a + b, 0) - cells;
+  while (over > 0) {
+    const i = widths.indexOf(Math.max(...widths));
+    if (widths[i]! <= 1) break;
+    widths[i]!--;
+    over--;
+  }
+  const drawn = widths.map((n, i) => fg(segments[i]!.color, (segments[i]!.glyph ?? FILL).repeat(n)));
+  return drawn.join("") + " ".repeat(Math.max(0, cells - widths.reduce((a, b) => a + b, 0)));
+}
+
+/**
+ * One cell of the campaign heatmap. With colour it is the GREENS ramp the
+ * README and the digest use; without it, the same five levels as shading.
+ */
+export function heatCell(attempts: number, body: string): string {
+  const level = Math.min(attempts, GREENS.length - 1);
+  return COLOR ? bg(GREENS[level]!, body) : [" ", "·", "▒", "▓", "█"][level]!.repeat(body.length);
+}
+
+/** One glyph per value, scaled to the largest. All-zero renders as a floor. */
+export function sparkline(values: number[], color = GREEN): string {
+  const peak = Math.max(...values, 0);
+  const line = values
+    .map((v) => (peak === 0 ? SPARKS[0]! : SPARKS[Math.min(7, Math.round((v / peak) * 7))]!))
+    .join("");
+  return fg(color, line);
+}
+
+// ── layout ───────────────────────────────────────────────────────
+
+/** A section rule that fills the width: `── name ────────────`. */
+export function rule(title: string, cells: number): string {
+  const label = ` ${title} `;
+  return (
+    fg(LINE, "──") + fg(MUTED, label) + fg(LINE, "─".repeat(Math.max(0, cells - 2 - label.length)))
+  );
+}
+
+// ── screen ───────────────────────────────────────────────────────
+
+const ALT_ON = "\x1b[?1049h\x1b[?25l";
+const ALT_OFF = "\x1b[?25h\x1b[?1049l";
+
+export interface Screen {
+  readonly rows: number;
+  readonly columns: number;
+  /** Draw a frame, clipped to the terminal height. */
+  draw(lines: string[]): void;
+  close(): void;
+}
+
+/**
+ * Take over the terminal: alternate buffer, hidden cursor, raw keys. `onKey`
+ * receives the raw chunk (`"\t"`, `"\x1b[B"`, `"q"`, …) — decoding belongs to
+ * the app, which is the only thing that knows what a key means.
+ *
+ * The caller MUST close() in a finally: raw mode survives a thrown exception.
+ */
+export function openScreen(onKey: (chunk: string) => void, onResize: () => void): Screen {
+  const out = process.stdout;
+  const input = process.stdin;
+
+  out.write(ALT_ON);
+  input.setRawMode?.(true);
+  input.setEncoding("utf8");
+  input.resume();
+  input.on("data", onKey);
+  out.on("resize", onResize);
+
+  let closed = false;
+  return {
+    get rows() {
+      return out.rows ?? 24;
+    },
+    get columns() {
+      return out.columns ?? 80;
+    },
+    draw(lines: string[]): void {
+      const visible = lines.slice(0, this.rows);
+      // Home, then erase each line as it is rewritten plus everything below the
+      // last one. No full clear, so a redraw does not flash, and no trailing
+      // newline, which would scroll the alternate screen by a row per frame.
+      out.write(`\x1b[H${visible.map((l) => `${l}\x1b[K`).join("\r\n")}\x1b[J`);
+    },
+    close(): void {
+      if (closed) return;
+      closed = true;
+      input.off("data", onKey);
+      out.off("resize", onResize);
+      input.setRawMode?.(false);
+      input.pause();
+      out.write(ALT_OFF);
+    },
+  };
+}

apps/cli/work.ts

@@ -0,0 +1,197 @@
+/**
+ * The work/ layout: one directory per spaced-repetition window.
+ *
+ *   work/1/<Difficulty>/<Category>/<num>.<slug>.<ext>   first solve
+ *   work/3/<Difficulty>/<Category>/<num>.<slug>.<ext>   3-day blind re-solve
+ *   work/7/<Difficulty>/<Category>/<num>.<slug>.<ext>   7-day blind re-solve
+ *
+ * The bucket name is the review window in days — the same +3/+7 windows the
+ * campaign's `Spaced Repetition — <date>` issues use (WINDOWS in
+ * spaced-repetition.ts). Below the bucket the layout is leetcode-cli's own
+ * `workDir/Difficulty/Category/` shape, so a re-solve lands beside its first
+ * solve under a different bucket.
+ *
+ * A window is *owed* when the first solve is at least that many days old and
+ * the bucket has no file for that problem yet. Reconciled from the filesystem
+ * on every run the way close-solved.ts reconciles issues: no state file, no
+ * bookkeeping, and a re-solve closes its window just by existing — presence,
+ * not content, so a stub in work/3 counts as done the same way `bun run pick`
+ * treats a scaffold under work/1 as picked. The first solve is the one place
+ * content is read: an empty scaffold in work/1 owes nothing, it is simply
+ * unsolved (isImplemented() in source.ts, the same gate close-solved.ts uses
+ * before it closes an issue).
+ *
+ * "First solve" is the ET date of the commit that first added the problem's
+ * work/ file — the same push that closes its problem issue and therefore
+ * starts the campaign's +3/+7 windows. An uncommitted solution has no date and
+ * is never owed a review yet: you just solved it. ET dates and the date
+ * arithmetic come from apps/api/src/srs.ts so the local picker and D1 cannot
+ * disagree about what day it is.
+ *
+ * Library: side-effect-free on import — nothing scans or shells out until a
+ * function is called.
+ */
+import { readFileSync } from "node:fs";
+import { join } from "node:path";
+import { WINDOWS, type Window, daysBetween } from "../api/src/srs.ts";
+import { isImplemented } from "./source.ts";
+
+export const ROOT = join(import.meta.dir, "..", "..");
+export const WORK = join(ROOT, "work");
+
+// The windows are the SRS ladder's rungs: srs.ts defines them once and the
+// bucket directories are named after them. Re-exported so the picker needs
+// only this module to know the work/ layout.
+export { WINDOWS, type Window };
+
+/** work/1 is the first solve, the rest are the review windows. */
+export const BUCKETS = [1, ...WINDOWS] as const;
+export type Bucket = (typeof BUCKETS)[number];
+const BUCKET_BY_DIR: Record<string, Bucket> = { "1": 1, "3": 3, "7": 7 };
+
+export function bucketDir(bucket: Bucket): string {
+  return join(WORK, String(bucket));
+}
+
+export interface WorkFile {
+  lc: number;
+  slug: string;
+  bucket: Bucket;
+  /** Path relative to work/, bucket included. */
+  rel: string;
+  difficulty: string;
+  category: string;
+}
+
+/** `<num>.<slug>.<ext>` — the only file name the campaign scripts parse. */
+function parseName(name: string): { lc: number; slug: string } | null {
+  const m = name.match(/^(\d+)\.(.+)\.[A-Za-z0-9]+$/);
+  return m ? { lc: Number(m[1]), slug: m[2]! } : null;
+}
+
+/** Parse `1/Medium/Array/15.3sum.py` — null for anything off-layout. */
+export function parseWorkPath(rel: string): WorkFile | null {
+  const parts = rel.split("/");
+  const name = parts.pop();
+  const bucket = BUCKET_BY_DIR[parts[0] ?? ""];
+  if (name === undefined || bucket === undefined) return null;
+  const parsed = parseName(name);
+  if (!parsed) return null;
+  return { ...parsed, bucket, rel, difficulty: parts[1] ?? "", category: parts[2] ?? "" };
+}
+
+/** LC number -> its files in that bucket (a problem may have both js and py). */
+export function scanBucket(bucket: Bucket): Map<number, WorkFile[]> {
+  const files = new Map<number, WorkFile[]>();
+  for (const name of new Bun.Glob("**/*.*").scanSync({ cwd: bucketDir(bucket) })) {
+    const file = parseWorkPath(`${bucket}/${name}`);
+    if (!file) continue;
+    const found = files.get(file.lc);
+    if (found) found.push(file);
+    else files.set(file.lc, [file]);
+  }
+  return files;
+}
+
+/**
+ * Is this file a real solution rather than an empty scaffold? The same call
+ * close-solved.ts makes before it closes an issue — a stub under work/1 is a
+ * problem still unsolved, and owes no review.
+ */
+export function isSolution(file: WorkFile): boolean {
+  const source = readFileSync(join(WORK, file.rel), "utf8");
+  return isImplemented(source, file.rel.endsWith(".py") ? "py" : "js");
+}
+
+/** Problems in a bucket with a solution on disk; scaffolds do not count. */
+export function solvedInBucket(bucket: Bucket): number {
+  let solved = 0;
+  for (const files of scanBucket(bucket).values()) {
+    if (files.some(isSolution)) solved++;
+  }
+  return solved;
+}
+
+/**
+ * LC number -> ET date of the commit that first added a work/ file for it.
+ *
+ * One `git log` over the whole of work/, oldest commit first, so the first
+ * date seen for a problem wins. Keyed by LC number rather than by path, which
+ * makes every past reshuffle of the directory layout — including the move into
+ * work/1 — irrelevant: the original add still counts. TZ + `format-local`
+ * put the dates on the ET calendar srs.ts speaks.
+ */
+export function firstSolved(): Map<number, string> {
+  const git = Bun.spawnSync(
+    [
+      "git",
+      "log",
+      "--reverse",
+      "--no-merges",
+      "--diff-filter=A",
+      "--date=format-local:%Y-%m-%d",
+      "--format=%ad",
+      "--name-only",
+      "--",
+      "work",
+    ],
+    { cwd: ROOT, env: { ...process.env, TZ: "America/New_York" } },
+  );
+  if (git.exitCode !== 0) {
+    throw new Error(`git log failed: ${git.stderr.toString().trim()}`);
+  }
+
+  const dates = new Map<number, string>();
+  let date = "";
+  for (const line of git.stdout.toString().split("\n")) {
+    if (/^\d{4}-\d{2}-\d{2}$/.test(line)) {
+      date = line;
+      continue;
+    }
+    const parsed = line === "" ? null : parseName(line.slice(line.lastIndexOf("/") + 1));
+    if (parsed && !dates.has(parsed.lc)) dates.set(parsed.lc, date);
+  }
+  return dates;
+}
+
+export interface Owed extends WorkFile {
+  /** ET date of the first solve. */
+  solved: string;
+  /** Days since the first solve — at least the window. */
+  days: number;
+}
+
+/**
+ * Both review windows for `today`, oldest solve first.
+ *
+ * Uncapped, unlike the review issue's WINDOW_CAP: the issue is one day's
+ * assignment, this is everything still owed. A window skipped for a day stays
+ * on the list until its file exists.
+ */
+export function reviewQueues(today: string): Map<Window, Owed[]> {
+  const dates = firstSolved();
+
+  // An empty scaffold under work/1 is a problem still unsolved, not one owing
+  // a review.
+  const solved: { file: WorkFile; date: string }[] = [];
+  for (const files of scanBucket(1).values()) {
+    const date = dates.get(files[0]!.lc);
+    if (date === undefined) continue; // uncommitted: you only just solved it
+    const file = files.find(isSolution);
+    if (file) solved.push({ file, date });
+  }
+
+  const queues = new Map<Window, Owed[]>();
+  for (const window of WINDOWS) {
+    const done = scanBucket(window);
+    const owed: Owed[] = [];
+    for (const { file, date } of solved) {
+      if (done.has(file.lc)) continue;
+      const days = daysBetween(date, today);
+      if (days >= window) owed.push({ ...file, solved: date, days });
+    }
+    owed.sort((a, b) => b.days - a.days || a.lc - b.lc);
+    queues.set(window, owed);
+  }
+  return queues;
+}

apps/docs/.gitignore

@@ -0,0 +1 @@
+.blume-verify/

apps/docs/blume.config.ts

@@ -7,11 +7,14 @@ export default defineConfig({
     site: "https://prdlk.github.io",
     base: "/leetcode",
   },
+  content: {
+    root: "content",
+  },
   navigation: {
     tabs: [
       {
-        label: "Solutions",
-        path: "/solutions",
+        label: "Reference",
+        path: "/reference",
       },
     ],
   },

apps/docs/content/(array)/1-two-sum.mdx

@@ -32,6 +32,20 @@ Can you come up with an algorithm that is less than O(n^2) time complexity?
 - `-10^9 <= target <= 10^9`
 - Only one valid answer exists.
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["twoSum(nums, target)"]) --> I["seen = {} — value to index"]
+  I --> L{"more (i, num) in nums?"}
+  L -- no --> E(["return []"])
+  L -- yes --> C["complement = target - num"]
+  C --> H{"complement in seen?"}
+  H -- yes --> R(["return [seen[complement], i]"])
+  H -- no --> W["seen[num] = i"]
+  W --> L
+```
+
 ## Solution
 
 ```py
@@ -45,3 +59,7 @@ class Solution:
             seen[num] = i
         return []
 ```
+
+## Explanation
+
+<YouTube id="KLlXCFG5TnA" title="Two Sum - Leetcode 1 - HashMap - Python" />

apps/docs/content/(array)/11-container-with-most-water.mdx

@@ -27,6 +27,22 @@ Notice that you may not slant the container.
 - `2 <= n <= 10^5`
 - `0 <= height[i] <= 10^4`
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["maxArea(height)"]) --> I["res = 0, left = 0, right = len - 1"]
+  I --> W{"left < right?"}
+  W -- no --> E(["return res"])
+  W -- yes --> A["area = (right - left) * min(height[left], height[right])"]
+  A --> M["res = max(res, area)"]
+  M --> C{"height[left] < height[right]?"}
+  C -- yes --> L["left += 1 — drop the shorter wall"]
+  C -- no --> R["right -= 1 — drop the shorter wall"]
+  L --> W
+  R --> W
+```
+
 ## Solution
 
 ```py
@@ -46,3 +62,7 @@ class Solution:
 
         return res
 ```
+
+## Explanation
+
+<YouTube id="UuiTKBwPgAo" title="Container with Most Water - Leetcode 11 - Python" />

apps/docs/content/(array)/121-best-time-to-buy-and-sell-stock.mdx

@@ -23,6 +23,24 @@ sidebar:
 - `1 <= prices.length <= 10^5`
 - `0 <= prices[i] <= 10^4`
 
+## Approach
+
+::::warning
+The diagram traces the code below literally. That code does not run: `curr` and `ans` are never initialised, so the first `curr > left` test raises `NameError`, and `left` holds a price rather than an index, so `prices[left]` indexes by value.
+::::
+
+```mermaid
+flowchart TD
+  S(["maxProfit(prices)"]) --> I["left = min(prices)"]
+  I --> F{"more right in range(len(prices))?"}
+  F -- no --> E(["return ans"])
+  F -- yes --> W{"curr > left?"}
+  W -- yes --> P["curr -= prices[left]; left += 1"]
+  P --> W
+  W -- no --> M["ans = max(ans, curr)"]
+  M --> F
+```
+
 ## Solution
 
 ```py
@@ -36,3 +54,7 @@ class Solution:
             ans = max(ans, curr)
         return ans
 ```
+
+## Explanation
+
+<YouTube id="1pkOgXD63yU" title="Sliding Window: Best Time to Buy and Sell Stock - Leetcode 121 - Python" />

apps/docs/content/(array)/1365-how-many-numbers-are-smaller-than-the-current-number.mdx

@@ -26,6 +26,20 @@ sidebar:
 - `2 <= nums.length <= 500`
 - `0 <= nums[i] <= 100`
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["smallerNumbersThanCurrent(nums)"]) --> F["freq: count every value in nums"]
+  F --> O["sorted = keys of freq, ascending"]
+  O --> I["count = 0, smaller = {}"]
+  I --> L{"more num in sorted?"}
+  L -- yes --> A["smaller[num] = count — everything already passed is smaller"]
+  A --> B["count += freq[num]"]
+  B --> L
+  L -- no --> M(["return nums.map(n => smaller[n])"])
+```
+
 ## Solution
 
 ```js

apps/docs/content/(array)/1413-minimum-value-to-get-positive-step-by-step-sum.mdx

@@ -37,6 +37,18 @@ startValue = 4 | startValue = 5 | nums
 - `1 <= nums.length <= 100`
 - `-100 <= nums[i] <= 100`
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["minStartValue(nums)"]) --> I["prefix = [nums[0]]"]
+  I --> L{"more i in 1..n-1?"}
+  L -- yes --> P["prefix.push(prefix[i-1] + nums[i])"]
+  P --> L
+  L -- no --> M["min = smallest running total"]
+  M --> R(["return max(1, 1 - min) — lift the dip to 1"])
+```
+
 ## Solution
 
 ```js

apps/docs/content/(array)/1426-counting-elements.mdx

@@ -23,6 +23,19 @@ sidebar:
 - `1 <= arr.length <= 1000`
 - `0 <= arr[i] <= 1000`
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["countElements(arr)"]) --> I["arrSet = Set(arr), count = 0"]
+  I --> L{"more n in arr?"}
+  L -- no --> E(["return count"])
+  L -- yes --> H{"arrSet has n + 1?"}
+  H -- yes --> C["count++"]
+  C --> L
+  H -- no --> L
+```
+
 ## Solution
 
 ```js

apps/docs/content/(array)/1480-running-sum-of-1d-array.mdx

@@ -27,6 +27,17 @@ sidebar:
 - `1 <= nums.length <= 1000`
 - `-10^6 <= nums[i] <= 10^6`
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["runningSum(nums)"]) --> I["prefix = [nums[0]]"]
+  I --> L{"more i in 1..n-1?"}
+  L -- yes --> P["prefix.push(prefix[i-1] + nums[i])"]
+  P --> L
+  L -- no --> E(["return prefix"])
+```
+
 ## Solution
 
 ```js

apps/docs/content/(array)/15-3sum.mdx

@@ -32,6 +32,28 @@ Notice that the solution set must not contain duplicate triplets.
 - `3 <= nums.length <= 3000`
 - `-10^5 <= nums[i] <= 10^5`
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["threeSum(nums)"]) --> O["nums.sort() — duplicates become adjacent"]
+  O --> F{"more i in 0..n-1?"}
+  F -- no --> E(["return result"])
+  F -- yes --> D{"nums[i] == nums[i-1]?"}
+  D -- yes --> F
+  D -- no --> P["left = i+1, right = n-1, target = -nums[i]"]
+  P --> W{"left < right?"}
+  W -- no --> F
+  W -- yes --> C["current = nums[left] + nums[right]"]
+  C --> Q{"current vs target"}
+  Q -- equal --> A["append triplet, skip equal neighbours, left += 1, right -= 1"]
+  A --> W
+  Q -- "current < target" --> L["left += 1 — need a bigger sum"]
+  L --> W
+  Q -- "current > target" --> R["right -= 1 — need a smaller sum"]
+  R --> W
+```
+
 ## Solution
 
 ```py
@@ -76,3 +98,7 @@ class Solution:
 
         return result
 ```
+
+## Explanation
+
+<YouTube id="jzZsG8n2R9A" title="3Sum - Leetcode 15 - Python" />

apps/docs/content/(array)/150-evaluate-reverse-polish-notation.mdx

@@ -0,0 +1,74 @@
+---
+title: '150. Evaluate Reverse Polish Notation'
+description: You are given an array of strings tokens that represents an arithmetic expression in a Reverse Polish Notation
+sidebar:
+  label: 'Evaluate Reverse Polish Notation'
+  badge: 'Medium'
+---
+
+<Badge variant="accent">Stack</Badge>
+
+### Example 1:
+- Input: `tokens = ["2","1","+","3","*"]`
+- Output: `9`
+- Explanation: `((2 + 1`) * `3`) = `9`
+
+### Example 2:
+- Input: `tokens = ["4","13","5","/","+"]`
+- Output: `6`
+- Explanation: `(4 + (13 / 5`)) = `6`
+
+### Example 3:
+- Input: `tokens = ["10","6","9","3","+","-11","*","/","*","17","+","5","+"]`
+- Output: `22`
+- Explanation: `((10 * (6 / ((9 + 3`) * `-11`))) + `17`) + `5 = ((10 * (6 / (12 * -11`))) + `17`) + `5 = ((10 * (6 / -132`)) + `17`) + `5 = ((10 * 0`) + `17`) + `5 = (0 + 17`) + `5 = 17 + 5 = 22`
+
+### Constraints:
+
+- `1 <= tokens.length <= 10^4`
+- `tokens[i]` is either an operator: "+", "-", "*", or "/", or an integer in the range [-200, 200].
+
+## Approach
+
+```mermaid
+flowchart TD
+  S(["evalRPN(tokens)"]) --> I["stack = [], operations = + - * /"]
+  I --> L{"more token c?"}
+  L -- no --> E(["return stack[0]"])
+  L -- yes --> Q{"c is an operator?"}
+  Q -- no --> N["stack.append(int(c))"]
+  N --> L
+  Q -- yes --> P["y = stack.pop(), x = stack.pop() — order matters"]
+  P --> C["stack.append(operations[c](x, y))"]
+  C --> L
+```
+
+## Solution
+
+```py
+class Solution:
+    def evalRPN(self, tokens: List[str]) -> int:
+        stack = []
+        operations = {
+            "+": lambda x, y: int(x + y),
+            "-": lambda x, y: int(x - y),
+            "*": lambda x, y: int(x * y),
+            "/": lambda x, y: int(x / y),
+        }
+
+        for c in tokens:
+            if c in operations:
+                y = stack.pop()
+                x = stack.pop()
+                calc_func = operations[c]
+                result = calc_func(x, y)
+                stack.append(result)
+            else:
+                stack.append(int(c))
+
+        return stack[0]
+```
+
+## Explanation
+
+<YouTube id="iu0082c4HDE" title="Evaluate Reverse Polish Notation - Leetcode 150 - Python" />

apps/docs/content/(array)/153-find-minimum-in-rotated-sorted-array.mdx

@@ -0,0 +1,74 @@
+---
+title: '153. Find Minimum in Rotated Sorted Array'
+description: 'Suppose an array of length n sorted in ascending order is rotated between 1 and n times. For example, the array nums = [0,1,2,4,5,6,7] might become:'
+sidebar:
+  label: 'Find Minimum in Rotated Sorted Array'
+  badge: 'Medium'
+---
+
+<Badge variant="accent">Binary Search</Badge>
+
+::::warning
+Notice that rotating an array [a[0], a[1], a[2], ..., a[n-1]] 1 time results in the array [a[n-1], a[0], a[1], a[2], ..., a[n-2]].
+::::
+
+### Example 1:
+- Input: `nums = [3,4,5,1,2]`
+- Output: `1`
+- Explanation: The original array was `[1,2,3,4,5]` rotated `3` times.
+
+### Example 2:
+- Input: `nums = [4,5,6,7,0,1,2]`
+- Output: `0`
+- Explanation: The original array was `[0,1,2,4,5,6,7]` and it was rotated `4` times.
+
+### Example 3:
+- Input: `nums = [11,13,15,17]`
+- Output: `11`
+- Explanation: The original array was `[11,13,15,17]` and it was rotated `4` times.
+
+### Constraints:
+
+- `n == nums.length`
+- `1 <= n <= 5000`
+- `-5000 <= nums[i] <= 5000`
+- All the integers of `nums` are unique.
+- `nums` is sorted and rotated between 1 and `n` times.
+
+## Approach
+
+```mermaid
+flowchart TD
+  S(["findMin(nums)"]) --> I["l = 0, r = n-1, lowest_index = -1"]
+  I --> W{"l <= r?"}
+  W -- no --> E(["return nums[lowest_index]"])
+  W -- yes --> M["m = (l + r) // 2"]
+  M --> C{"nums[m] <= nums[-1]?"}
+  C -- yes --> A["in the right sorted run: record lowest_index = m, r = m - 1"]
+  A --> W
+  C -- no --> B["still in the left sorted run: l = m + 1"]
+  B --> W
+```
+
+## Solution
+
+```py
+class Solution:
+    def findMin(self, nums: List[int]) -> int:
+        l, r = 0, len(nums) - 1
+        lowest_index = -1
+
+        while l <= r:
+            m = (l + r) // 2
+            if nums[m] <= nums[-1]:
+                lowest_index = m
+                r = m - 1
+            else:
+                l = m + 1
+
+        return nums[lowest_index]
+```
+
+## Explanation
+
+<YouTube id="nIVW4P8b1VA" title="Find Minimum in Rotated Sorted Array - Binary Search - Leetcode 153 - Python" />

apps/docs/content/(array)/1636-sort-array-by-increasing-frequency.mdx

@@ -27,6 +27,18 @@ sidebar:
 - `1 <= nums.length <= 100`
 - `-100 <= nums[i] <= 100`
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["frequencySort(nums)"]) --> F["freq: count every value in nums"]
+  F --> C["sort nums with a two-key comparator"]
+  C --> K1["primary: freq[a] - freq[b] — rarer first"]
+  C --> K2["tie-break: b - a — larger value first"]
+  K1 --> R(["return nums"])
+  K2 --> R
+```
+
 ## Solution
 
 ```js
@@ -40,3 +52,7 @@ var frequencySort = function (nums) {
   return nums.sort((a, b) => freq[a] - freq[b] || b - a);
 };
 ```
+
+## Explanation
+
+<YouTube id="Evq1SfUbhBg" title="Sort Array by Increasing Frequency - Leetcode 1636 - Python" />

apps/docs/content/(array)/167-two-sum-ii-input-array-is-sorted.mdx

@@ -35,6 +35,22 @@ Your solution must use only constant extra space.
 - `-1000 <= target <= 1000`
 - The tests are generated such that there is exactly one solution.
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["twoSum(numbers, target)"]) --> I["i = 0, j = len - 1"]
+  I --> W{"i < j?"}
+  W -- no --> E(["return []"])
+  W -- yes --> C["c = numbers[i] + numbers[j]"]
+  C --> Q{"c vs target"}
+  Q -- equal --> R(["return [i+1, j+1] — 1-indexed"])
+  Q -- "c < target" --> L["i += 1 — grow the sum"]
+  L --> W
+  Q -- "c > target" --> H["j -= 1 — shrink the sum"]
+  H --> W
+```
+
 ## Solution
 
 ```py
@@ -53,3 +69,7 @@ class Solution:
                 j-=1
         return []
 ```
+
+## Explanation
+
+<YouTube id="cQ1Oz4ckceM" title="TWO SUM II - Amazon Coding Interview Question - Leetcode 167 - Python" />

apps/docs/content/(array)/217-contains-duplicate.mdx

@@ -27,6 +27,19 @@ sidebar:
 - `1 <= nums.length <= 10^5`
 - `-10^9 <= nums[i] <= 10^9`
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["containsDuplicate(nums)"]) --> I["freq = {}"]
+  I --> L{"more n in nums?"}
+  L -- no --> E(["return false"])
+  L -- yes --> C["freq[n] = (freq[n] or 0) + 1"]
+  C --> Q{"freq[n] >= 2?"}
+  Q -- yes --> R(["return true"])
+  Q -- no --> L
+```
+
 ## Solution
 
 ```js
@@ -45,3 +58,7 @@ var containsDuplicate = function(nums) {
   return false;
 };
 ```
+
+## Explanation
+
+<YouTube id="3OamzN90kPg" title="Contains Duplicate - Leetcode 217 - Python" />

apps/docs/content/(array)/238-product-of-array-except-self.mdx

@@ -27,6 +27,17 @@ Can you solve the problem in O(1) extra space complexity? (The output array does
 - `-30 <= nums[i] <= 30`
 - The input is generated such that `answer[i]` is guaranteed to fit in a 32-bit integer.
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["productExceptSelf(nums)"]) --> I["answer = ones(n), rightArr = ones(n)"]
+  I --> P1["Pass 1 — left to right: answer[i] = nums[i-1] * answer[i-1]"]
+  P1 --> P2["Pass 2 — right to left: rightArr[i] = nums[i+1] * rightArr[i+1]"]
+  P2 --> P3["Pass 3: answer[i] *= rightArr[i]"]
+  P3 --> E(["return answer — no division needed"])
+```
+
 ## Solution
 
 ```js
@@ -51,3 +62,7 @@ var productExceptSelf = function(nums) {
   return answer;
 };
 ```
+
+## Explanation
+
+<YouTube id="bNvIQI2wAjk" title="Product of Array Except Self - Leetcode 238 - Python" />

apps/docs/content/(array)/268-missing-number.mdx

@@ -34,6 +34,18 @@ Could you implement a solution using only O(1) extra space complexity and O(n) r
 - `0 <= nums[i] <= n`
 - All the numbers of `nums` are unique.
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["missingNumber(nums)"]) --> I["numSet = Set(nums), expectedCount = n + 1"]
+  I --> L{"more i in 0..expectedCount-1?"}
+  L -- yes --> Q{"numSet has i?"}
+  Q -- no --> R(["return i"])
+  Q -- yes --> L
+  L -- no --> E(["return -1"])
+```
+
 ## Solution
 
 ```js
@@ -53,3 +65,7 @@ var missingNumber = function(nums) {
   return -1;
 };
 ```
+
+## Explanation
+
+<YouTube id="WnPLSRLSANE" title="Missing Number - Blind 75 - Leetcode 268 - Python" />

apps/docs/content/(array)/303-range-sum-query-immutable.mdx

@@ -20,6 +20,21 @@ sidebar:
 - `0 <= left <= right < nums.length`
 - At most `10^4` calls will be made to `sumRange`.
 
+## Approach
+
+```mermaid
+flowchart TD
+  subgraph build["constructor(nums) — O(n) once"]
+    A["prefix = [0]"] --> B{"more n in nums?"}
+    B -- yes --> C["prefix.push(last + n)"]
+    C --> B
+  end
+  build --> Q
+  subgraph Q["sumRange(left, right) — O(1) per call"]
+    D(["return prefix[right + 1] - prefix[left]"])
+  end
+```
+
 ## Solution
 
 ```js
@@ -50,3 +65,7 @@ class NumArray {
  * var param_1 = obj.sumRange(left,right)
  */
 ```
+
+## Explanation
+
+<YouTube id="2pndAmo_sMA" title="Range Sum Query Immutable - Leetcode 303 - Python" />

apps/docs/content/(array)/33-search-in-rotated-sorted-array.mdx

@@ -32,6 +32,29 @@ You must write an algorithm with O(log n) runtime complexity.
 - `nums` is an ascending array that is possibly rotated.
 - `-10^4 <= target <= 10^4`
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["search(nums, target)"]) --> I["l = 0, r = n - 1"]
+  I --> W{"l <= r?"}
+  W -- no --> E(["return -1"])
+  W -- yes --> M["mid = floor((l + r) / 2)"]
+  M --> F{"nums[mid] == target?"}
+  F -- yes --> R(["return mid"])
+  F -- no --> H{"nums[l] <= nums[mid]? — left half sorted"}
+  H -- yes --> A{"target outside [nums[l], nums[mid]]?"}
+  A -- yes --> A1["l = mid + 1 — search the right half"]
+  A -- no --> A2["r = mid - 1"]
+  H -- no --> B{"target outside [nums[mid], nums[r]]?"}
+  B -- yes --> B1["r = mid - 1 — search the left half"]
+  B -- no --> B2["l = mid + 1"]
+  A1 --> W
+  A2 --> W
+  B1 --> W
+  B2 --> W
+```
+
 ## Solution
 
 ```js
@@ -69,3 +92,7 @@ var search = function(nums, target) {
   return -1;
 };
 ```
+
+## Explanation
+
+<YouTube id="U8XENwh8Oy8" title="Search in rotated sorted array - Leetcode 33 - Python" />

apps/docs/content/(array)/347-top-k-frequent-elements.mdx

@@ -31,6 +31,16 @@ Your algorithm's time complexity must be better than O(n log n), where n is the
 - `k` is in the range `[1, the number of unique elements in the array]`.
 - It is guaranteed that the answer is unique.
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["topKFrequent(nums, k)"]) --> F["freq: count every value in nums"]
+  F --> K["keys of freq, cast to Number"]
+  K --> O["sort by freq[b] - freq[a] — most frequent first"]
+  O --> R(["return slice(0, k)"])
+```
+
 ## Solution
 
 ```js
@@ -49,3 +59,7 @@ var topKFrequent = function (nums, k) {
     .slice(0, k);
 };
 ```
+
+## Explanation
+
+<YouTube id="YPTqKIgVk-k" title="Top K Frequent Elements - Bucket Sort - Leetcode 347 - Python" />

apps/docs/content/(array)/35-search-insert-position.mdx

@@ -0,0 +1,71 @@
+---
+title: '35. Search Insert Position'
+description: Given a sorted array of distinct integers and a target value, return the index if the target is found. If not, return the index where it would be if it were inserted in order
+sidebar:
+  label: 'Search Insert Position'
+  badge: 'Easy'
+---
+
+<Badge variant="accent">Binary Search</Badge>
+
+::::warning
+You must write an algorithm with O(log n) runtime complexity.
+::::
+
+### Example 1:
+- Input: `nums = [1,3,5,6], target = 5`
+- Output: `2`
+
+### Example 2:
+- Input: `nums = [1,3,5,6], target = 2`
+- Output: `1`
+
+### Example 3:
+- Input: `nums = [1,3,5,6], target = 7`
+- Output: `4`
+
+### Constraints:
+
+- `1 <= nums.length <= 10^4`
+- `-10^4 <= nums[i] <= 10^4`
+- `nums` contains distinct values sorted in ascending order.
+- `-10^4 <= target <= 10^4`
+
+## Approach
+
+```mermaid
+flowchart TD
+  S(["searchInsert(nums, target)"]) --> I["l = 0, r = n - 1"]
+  I --> W{"l <= r?"}
+  W -- yes --> M["m = (l + r) // 2"]
+  M --> Q{"nums[m] vs target"}
+  Q -- equal --> R(["return m"])
+  Q -- "nums[m] < target" --> L["l = m + 1"]
+  L --> W
+  Q -- "nums[m] > target" --> H["r = m - 1"]
+  H --> W
+  W -- no --> E(["return (l + r) // 2 + 1 — the insert slot"])
+```
+
+## Solution
+
+```py
+class Solution:
+    def searchInsert(self, nums: List[int], target: int) -> int:
+        l, r = 0, len(nums) - 1
+
+        while l <= r:
+            m = (l + r) // 2
+            if nums[m] == target:
+                return m
+            elif nums[m] < target:
+                l = m + 1
+            elif nums[m] > target:
+                r = m - 1
+
+        return (l + r) // 2 + 1
+```
+
+## Explanation
+
+<YouTube id="K-RYzDZkzCI" title="Search Insert Position - Binary Search - Leetcode 35 - Python" />

apps/docs/content/(array)/48-rotate-image.mdx

@@ -0,0 +1,37 @@
+---
+title: '48. Rotate Image'
+description: You are given an n x n 2D matrix representing an image, rotate the image by 90 degrees (clockwise)
+sidebar:
+  label: 'Rotate Image'
+  badge: 'Medium'
+---
+
+<Badge variant="accent">Matrix Index Math</Badge>
+
+### Example 1:
+- Input: `matrix = [[1,2,3],[4,5,6],[7,8,9]]`
+- Output: `[[7,4,1],[8,5,2],[9,6,3]]`
+
+### Example 2:
+- Input: `matrix = [[5,1,9,11],[2,4,8,10],[13,3,6,7],[15,14,12,16]]`
+- Output: `[[15,13,2,5],[14,3,4,1],[12,6,8,9],[16,7,10,11]]`
+
+### Constraints:
+
+- `n == matrix.length == matrix[i].length`
+- `1 <= n <= 20`
+- `-1000 <= matrix[i][j] <= 1000`
+
+## Solution
+
+```py
+class Solution:
+    def rotate(self, matrix: List[List[int]]) -> None:
+        n = len(matrix)
+        for r in range(n):
+            for c in range(r + 1, n):
+                matrix[r][c], matrix[c][r] = matrix[c][r], matrix[r][c]
+
+        for r in range(n):
+            matrix[r].reverse()
+```

apps/docs/content/(array)/49-group-anagrams.mdx

@@ -27,6 +27,21 @@ sidebar:
 - `0 <= strs[i].length <= 100`
 - `strs[i]` consists of lowercase English letters.
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["groupAnagrams(strs)"]) --> K["sorted[i] = letters of strs[i], sorted — the anagram key"]
+  K --> I["anagrams = {}"]
+  I --> L{"more i in 0..n-1?"}
+  L -- no --> E(["return Object.values(anagrams)"])
+  L -- yes --> Q{"anagrams has sorted[i]?"}
+  Q -- no --> N["anagrams[sorted[i]] = [strs[i]]"]
+  Q -- yes --> P["anagrams[sorted[i]].push(strs[i])"]
+  N --> L
+  P --> L
+```
+
 ## Solution
 
 ```js
@@ -49,3 +64,7 @@ var groupAnagrams = function(strs) {
   return Object.values(anagrams);
 };
 ```
+
+## Explanation
+
+<YouTube id="vzdNOK2oB2E" title="Group Anagrams - Categorize Strings by Count - Leetcode 49" />

apps/docs/content/(array)/54-spiral-matrix.mdx

@@ -0,0 +1,59 @@
+---
+title: '54. Spiral Matrix'
+description: Given an m x n matrix, return all elements of the matrix in spiral order
+sidebar:
+  label: 'Spiral Matrix'
+  badge: 'Medium'
+---
+
+<Badge variant="accent">Matrix Index Math</Badge>
+
+### Example 1:
+- Input: `matrix = [[1,2,3],[4,5,6],[7,8,9]]`
+- Output: `[1,2,3,6,9,8,7,4,5]`
+
+### Example 2:
+- Input: `matrix = [[1,2,3,4],[5,6,7,8],[9,10,11,12]]`
+- Output: `[1,2,3,4,8,12,11,10,9,5,6,7]`
+
+### Constraints:
+
+- `m == matrix.length`
+- `n == matrix[i].length`
+- `1 <= m, n <= 10`
+- `-100 <= matrix[i][j] <= 100`
+
+## Solution
+
+```py
+class Solution:
+    def spiralOrder(self, matrix: List[List[int]]) -> List[int]:
+        res = []
+        top, bottom = 0, len(matrix) - 1
+        left, right = 0, len(matrix[0]) - 1
+
+        while top <= bottom and left <= right:
+            # 1. Top Row
+            for c in range(left, right + 1):
+                res.append(matrix[top][c])
+            top += 1
+
+            # 2. Right Column
+            for r in range(top, bottom + 1):
+                res.append(matrix[r][right])
+            right -= 1
+
+            # 3. Bottom Row
+            if top <= bottom:
+                for c in range(right, left - 1, -1):
+                    res.append(matrix[bottom][c])
+                bottom -= 1
+
+            # 4. Left Column
+            if left <= right:
+                for r in range(bottom, top - 1, -1):
+                    res.append(matrix[r][left])
+                left += 1
+
+        return res
+```

apps/docs/content/(array)/560-subarray-sum-equals-k.mdx

@@ -22,6 +22,21 @@ sidebar:
 - `-1000 <= nums[i] <= 1000`
 - `-10^7 <= k <= 10^7`
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["subarraySum(nums, k)"]) --> I["map = {0: 1} — the empty prefix; sum = 0, count = 0"]
+  I --> L{"more n in nums?"}
+  L -- no --> E(["return count"])
+  L -- yes --> A["sum += n"]
+  A --> Q{"map has sum - k?"}
+  Q -- yes --> C["count += map.get(sum - k) — every earlier prefix that closes a window"]
+  Q -- no --> W["map.set(sum, count of sum + 1)"]
+  C --> W
+  W --> L
+```
+
 ## Solution
 
 ```js
@@ -49,3 +64,7 @@ var subarraySum = function(nums, k) {
   return count
 };
 ```
+
+## Explanation
+
+<YouTube id="fFVZt-6sgyo" title="Subarray Sum Equals K - Prefix Sums - Leetcode 560 - Python" />

apps/docs/content/(array)/704-binary-search.mdx

@@ -29,6 +29,22 @@ You must write an algorithm with O(log n) runtime complexity.
 - All the integers in `nums` are unique.
 - `nums` is sorted in ascending order.
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["search(nums, target)"]) --> I["left = 0, right = n - 1"]
+  I --> W{"left <= right?"}
+  W -- no --> E(["return -1"])
+  W -- yes --> M["mid = left + floor((right - left) / 2)"]
+  M --> Q{"nums[mid] vs target"}
+  Q -- equal --> R(["return mid"])
+  Q -- "nums[mid] < target" --> L["left = mid + 1"]
+  L --> W
+  Q -- "nums[mid] > target" --> H["right = mid - 1"]
+  H --> W
+```
+
 ## Solution
 
 ```js
@@ -54,3 +70,7 @@ var search = function(nums, target) {
     return -1;
 };
 ```
+
+## Explanation
+
+<YouTube id="s4DPM8ct1pI" title="Binary Search - Leetcode 704 - Python" />

apps/docs/content/(array)/73-set-matrix-zeroes.mdx

@@ -0,0 +1,67 @@
+---
+title: '73. Set Matrix Zeroes'
+description: Given an m x n integer matrix matrix, if an element is 0, set its entire row and column to 0's
+sidebar:
+  label: 'Set Matrix Zeroes'
+  badge: 'Medium'
+---
+
+<Badge variant="accent">Matrix Index Math</Badge>
+
+::::warning
+You must do it in place.
+::::
+
+### Example 1:
+- Input: `matrix = [[1,1,1],[1,0,1],[1,1,1]]`
+- Output: `[[1,0,1],[0,0,0],[1,0,1]]`
+
+### Example 2:
+- Input: `matrix = [[0,1,2,0],[3,4,5,2],[1,3,1,5]]`
+- Output: `[[0,0,0,0],[0,4,5,0],[0,3,1,0]]`
+
+### Constraints:
+
+- `m == matrix.length`
+- `n == matrix[0].length`
+- `1 <= m, n <= 200`
+- `-2^31 <= matrix[i][j] <= 2^31 - 1`
+- A straightforward solution using O(mn) space is probably a bad idea.
+- A simple improvement uses O(m + n) space, but still not the best solution.
+- Could you devise a constant space solution?
+
+## Solution
+
+```py
+class Solution:
+    def setZeroes(self, matrix: List[List[int]]) -> None:
+        """
+        Do not return anything, modify matrix in-place instead.
+        """
+        m, n = len(matrix), len(matrix[0])
+
+        # 1. remember if row 0 or column 0 have their own zeros
+        row0_zero = any(matrix[0][c] == 0 for c in range(n))
+        col0_zero = any(matrix[r][0] == 0 for r in range(m))
+
+        # 2. use row 0 and column 0 as marker lists
+        for r in range(1, m):
+            for c in range(1, n):
+                if matrix[r][c] == 0:
+                    matrix[r][0] = 0
+                    matrix[0][c] = 0
+
+        # 3. zero the inner cells from the markers
+        for r in range(1, m):
+            for c in range(1, n):
+                if matrix[r][0] == 0 or matrix[0][c] == 0:
+                    matrix[r][c] = 0
+
+        # 4. handle row 0 and column 0 last
+        if row0_zero:
+            for c in range(n):
+                matrix[0][c] = 0
+        if col0_zero:
+            for r in range(m):
+                matrix[r][0] = 0
+```

apps/docs/content/(array)/739-daily-temperatures.mdx

@@ -0,0 +1,59 @@
+---
+title: '739. Daily Temperatures'
+description: Given an array of integers temperatures represents the daily temperatures, return an array answer such that answer[i] is the number of days you have to wait after the i^th day to get a warmer temperature. If there is no future day for which this is possible, keep answer[i] == 0 instead
+sidebar:
+  label: 'Daily Temperatures'
+  badge: 'Medium'
+---
+
+<Badge variant="accent">Monotonic Stack</Badge>
+
+### Example 1:
+- Input: `temperatures = [73,74,75,71,69,72,76,73]`
+- Output: `[1,1,4,2,1,1,0,0]`
+
+### Example 2:
+- Input: `temperatures = [30,40,50,60]`
+- Output: `[1,1,1,0]`
+
+### Example 3:
+- Input: `temperatures = [30,60,90]`
+- Output: `[1,1,0]`
+
+### Constraints:
+
+- `1 <= temperatures.length <= 10^5`
+- `30 <= temperatures[i] <= 100`
+
+## Approach
+
+```mermaid
+flowchart TD
+  S(["dailyTemperatures(temperatures)"]) --> I["ans = zeros(n), stack = [] — indices of days still waiting"]
+  I --> L{"more (i, temp)?"}
+  L -- no --> E(["return ans"])
+  L -- yes --> W{"stack non-empty and temperatures[stack[-1]] < temp?"}
+  W -- yes --> P["prev = stack.pop(); ans[prev] = i - prev — today resolves that day"]
+  P --> W
+  W -- no --> A["stack.append(i)"]
+  A --> L
+```
+
+## Solution
+
+```py
+class Solution:
+    def dailyTemperatures(self, temperatures: List[int]) -> List[int]:
+        ans = [0] * len(temperatures)
+        stack = []
+        for i, temp in enumerate(temperatures):
+            while stack and temperatures[stack[-1]] < temp:
+                prev = stack.pop()
+                ans[prev] = i - prev
+            stack.append(i)
+        return ans
+```
+
+## Explanation
+
+<YouTube id="cTBiBSnjO3c" title="Daily Temperatures - Monotonic Stack - Leetcode 739 - Python" />

apps/docs/content/(array)/74-search-a-2d-matrix.mdx

@@ -0,0 +1,91 @@
+---
+title: '74. Search a 2D Matrix'
+description: 'You are given an m x n integer matrix matrix with the following two properties:'
+sidebar:
+  label: 'Search a 2D Matrix'
+  badge: 'Medium'
+---
+
+<Badge variant="accent">Binary Search</Badge>
+
+::::warning
+You must write a solution in O(log(m * n)) time complexity.
+::::
+
+### Example 1:
+- Input: `matrix = [[1,3,5,7],[10,11,16,20],[23,30,34,60]], target = 3`
+- Output: `true`
+
+### Example 2:
+- Input: `matrix = [[1,3,5,7],[10,11,16,20],[23,30,34,60]], target = 13`
+- Output: `false`
+
+### Constraints:
+
+- `m == matrix.length`
+- `n == matrix[i].length`
+- `1 <= m, n <= 100`
+- `-10^4 <= matrix[i][j], target <= 10^4`
+
+## Approach
+
+```mermaid
+flowchart TD
+  S(["searchMatrix(matrix, target)"]) --> I["top = 0, bot = ROWS - 1"]
+  I --> W1{"top <= bot?"}
+  W1 -- yes --> R1["row = (top + bot) // 2"]
+  R1 --> C1{"target vs that row's range"}
+  C1 -- "above matrix[row][-1]" --> U["top = row + 1"]
+  C1 -- "below matrix[row][0]" --> V["bot = row - 1"]
+  C1 -- inside --> B(["break — row found"])
+  U --> W1
+  V --> W1
+  W1 -- no --> F(["return False — no row can hold it"])
+  B --> P["row = (top + bot) // 2; l = 0, r = COLS - 1"]
+  P --> W2{"l <= r?"}
+  W2 -- no --> G(["return False"])
+  W2 -- yes --> M["m = (l + r) // 2"]
+  M --> C2{"target vs matrix[row][m]"}
+  C2 -- greater --> X["l = m + 1"]
+  C2 -- smaller --> Y["r = m - 1"]
+  C2 -- equal --> T(["return True"])
+  X --> W2
+  Y --> W2
+```
+
+## Solution
+
+```py
+class Solution:
+    def searchMatrix(self, matrix: List[List[int]], target: int) -> bool:
+        ROWS, COLS = len(matrix), len(matrix[0])
+
+        top, bot = 0, ROWS - 1
+        while top <= bot:
+            row = (top + bot) // 2
+            if target > matrix[row][-1]:
+                top = row + 1
+            elif target < matrix[row][0]:
+                bot = row - 1
+            else:
+                break
+
+        if not (top <= bot):
+            return False
+
+        row = (top + bot) // 2
+        l, r = 0, COLS - 1
+        while l <= r:
+            m = (l + r) // 2
+            if target > matrix[row][m]:
+                l = m + 1
+            elif target < matrix[row][m]:
+                r = m - 1
+            else:
+                return True
+        return False
+```
+
+## Explanation
+
+<YouTube id="Ber2pi2C0j0" title="Search a 2D Matrix - Leetcode 74 - Python" />

apps/docs/content/(array)/853-car-fleet.mdx

@@ -0,0 +1,67 @@
+---
+title: '853. Car Fleet'
+description: There are n cars at given miles away from the starting mile 0, traveling to reach the mile target
+sidebar:
+  label: 'Car Fleet'
+  badge: 'Medium'
+---
+
+<Badge variant="accent">Monotonic Stack</Badge>
+
+### Example 1:
+- Input: `target = 12, position = [10,8,0,5,3], speed = [2,4,1,1,3]`
+- Output: `3`
+- Explanation: • The cars starting at `10 (speed 2`) and `8 (speed 4`) become a fleet, meeting each other at `12`. The fleet forms at `target`. • The car starting at `0 (speed 1`) does not catch up to any other car, so it is a fleet by itself. • The cars starting at `5 (speed 1`) and `3 (speed 3`) become a fleet, meeting each other at `6`. The fleet moves at `speed 1` until it reaches `target`.
+
+### Example 2:
+- Input: `target = 10, position = [3], speed = [3]`
+- Output: `1`
+- Explanation: There is only one car, hence there is only one fleet.
+
+### Example 3:
+- Input: `target = 100, position = [0,2,4], speed = [4,2,1]`
+- Output: `1`
+- Explanation: • The cars starting at `0 (speed 4`) and `2 (speed 2`) become a fleet, meeting each other at `4`. The car starting at `4 (speed 1`) travels to `5`. • Then, the fleet at `4 (speed 2`) and the car at `position 5 (speed 1`) become one fleet, meeting each other at `6`. The fleet moves at `speed 1` until it reaches `target`.
+
+### Constraints:
+
+- `n == position.length == speed.length`
+- `1 <= n <= 10^5`
+- `0 < target <= 10^6`
+- `0 <= position[i] < target`
+- All the values of `position` are unique.
+- `0 < speed[i] <= 10^6`
+
+## Approach
+
+```mermaid
+flowchart TD
+  S(["carFleet(target, position, speed)"]) --> O["zip and sort by position, descending — nearest the target first"]
+  O --> I["stack = [] — arrival time of each fleet's leader"]
+  I --> L{"more (pos, spd)?"}
+  L -- no --> E(["return len(stack) — one entry per fleet"])
+  L -- yes --> T["time = (target - pos) / spd"]
+  T --> Q{"stack empty or time > stack[-1]?"}
+  Q -- yes --> A["stack.append(time) — slower, so it starts a new fleet"]
+  Q -- no --> B["discard — it catches the car ahead and joins that fleet"]
+  A --> L
+  B --> L
+```
+
+## Solution
+
+```py
+class Solution:
+    def carFleet(self, target: int, position: List[int], speed: List[int]) -> int:
+        stack = []
+        for pos, spd in sorted(zip(position, speed), reverse=True):
+            time = (target - pos) / spd
+            if not stack or time > stack[-1]:
+                stack.append(time)
+
+        return len(stack)
+```
+
+## Explanation
+
+<YouTube id="Pr6T-3yB9RM" title="Car Fleet - Leetcode 853 - Python" />

apps/docs/content/(array)/875-koko-eating-bananas.mdx

@@ -26,6 +26,22 @@ sidebar:
 - `piles.length <= h <= 10^9`
 - `1 <= piles[i] <= 10^9`
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["minEatingSpeed(piles, h)"]) --> I["l = 1, r = max(piles)"]
+  I --> W{"l < r?"}
+  W -- no --> E(["return l — the smallest workable speed"])
+  W -- yes --> M["mid = floor((l + r) / 2)"]
+  M --> K["kWorks(mid): hours = sum of ceil(p / mid)"]
+  K --> Q{"hours <= h?"}
+  Q -- yes --> A["r = mid — keep mid, try slower"]
+  Q -- no --> B["l = mid + 1 — too slow, speed up"]
+  A --> W
+  B --> W
+```
+
 ## Solution
 
 ```js
@@ -60,3 +76,7 @@ var minEatingSpeed = function(piles, h) {
   return l;
 };
 ```
+
+## Explanation
+
+<YouTube id="U2SozAs9RzA" title="Koko Eating Bananas - Binary Search - Leetcode 875 - Python" />

apps/docs/content/(array)/977-squares-of-a-sorted-array.mdx

@@ -27,6 +27,21 @@ Squaring each element and sorting the new array is very trivial, could you find
 - `-10^4 <= nums[i] <= 10^4`
 - `nums` is sorted in non-decreasing order.
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["sortedSquares(nums)"]) --> I["ans = Array(n), left = 0, right = n - 1"]
+  I --> L{"more i from n-1 down to 0?"}
+  L -- no --> E(["return ans"])
+  L -- yes --> C{"abs(nums[left]) < abs(nums[right])?"}
+  C -- yes --> R["square = nums[right]; right--"]
+  C -- no --> F["square = nums[left]; left++"]
+  R --> W["ans[i] = square * square — fill from the back, largest first"]
+  F --> W
+  W --> L
+```
+
 ## Solution
 
 ```js
@@ -53,3 +68,7 @@ var sortedSquares = function(nums) {
   return ans;
 };
 ```
+
+## Explanation
+
+<YouTube id="FPCZsG_AkUg" title="Squares of a Sorted Array - Leetcode 977 - Python" />

apps/docs/content/(hash-table)/141-linked-list-cycle.mdx

@@ -0,0 +1,74 @@
+---
+title: '141. Linked List Cycle'
+description: Given head, the head of a linked list, determine if the linked list has a cycle in it
+sidebar:
+  label: 'Linked List Cycle'
+  badge: 'Easy'
+---
+
+<Badge variant="accent">Linked List</Badge>
+
+::::warning
+Can you solve it using O(1) (i.e. constant) memory?
+::::
+
+### Example 1:
+- Input: `head = [3,2,0,-4], pos = 1`
+- Output: `true`
+- Explanation: There is a cycle in the linked list, where the tail connects to the 1st node (0-indexed).
+
+### Example 2:
+- Input: `head = [1,2], pos = 0`
+- Output: `true`
+- Explanation: There is a cycle in the linked list, where the tail connects to the 0th node.
+
+### Example 3:
+- Input: `head = [1], pos = -1`
+- Output: `false`
+- Explanation: There is no cycle in the linked list.
+
+### Constraints:
+
+- The number of the nodes in the list is in the range [0, 10^4].
+- `-10^5 <= Node.val <= 10^5`
+- pos is -1 or a valid index in the linked-list.
+
+## Approach
+
+```mermaid
+flowchart TD
+  S(["hasCycle(head)"]) --> I["fast = slow = head"]
+  I --> W{"fast and fast.next?"}
+  W -- no --> E(["return False — ran off the end, no cycle"])
+  W -- yes --> A["fast = fast.next.next — two steps"]
+  A --> B["slow = slow.next — one step"]
+  B --> Q{"fast is slow?"}
+  Q -- yes --> R(["return True — the gap closed, so it loops"])
+  Q -- no --> W
+```
+
+## Solution
+
+```py
+# Definition for singly-linked list.
+# class ListNode:
+#     def __init__(self, x):
+#         self.val = x
+#         self.next = None
+
+
+class Solution:
+    def hasCycle(self, head: Optional[ListNode]) -> bool:
+        fast = slow = head
+
+        while fast and fast.next:
+            fast = fast.next.next
+            slow = slow.next
+            if fast is slow:
+                return True
+        return False
+```
+
+## Explanation
+
+<YouTube id="gBTe7lFR3vc" title="Linked List Cycle - Floyd's Tortoise and Hare - Leetcode 141 - Python" />

apps/docs/content/(hash-table)/146-lru-cache.mdx

@@ -0,0 +1,115 @@
+---
+title: '146. LRU Cache'
+description: Design a data structure that follows the constraints of a Least Recently Used (LRU) cache
+sidebar:
+  label: 'LRU Cache'
+  badge: 'Medium'
+---
+
+<Badge variant="accent">Hybrid Structures</Badge>
+
+### Example 1:
+- Input: ``
+- Output: ``
+- Explanation: Input ["LRUCache", "put", "put", "get", "put", "get", "put", "get", "get", "get"] `[[2]`, [1, 1], [2, 2], `[1]`, [3, 3], `[2]`, [4, 4], `[1]`, `[3]`, [4]] Output [null, null, null, `1`, null, `-1`, null, `-1`, `3`, 4] Explanation LRUCache lRUCache = new LRUCache(2); lRUCache.put(1, `1`); // cache is {1=1} lRUCache.put(2, `2`); // cache is {1=1, 2=2} lRUCache.get(1);    // return `1` lRUCache.put(3, `3`); // LRU `key` was `2`, evicts `key 2`, cache is {1=1, 3=3} lRUCache.get(2);    // returns `-1` (not found) lRUCache.put(4, `4`); // LRU `key` was `1`, evicts `key 1`, cache is {4=4, 3=3} lRUCache.get(1);    // return `-1` (not found) lRUCache.get(3);    // return `3` lRUCache.get(4);    // return `4`
+
+### Constraints:
+
+- `1 <= capacity <= 3000`
+- `0 <= key <= 10^4`
+- `0 <= value <= 10^5`
+- At most 2 * 10^5 calls will be made to get and put.
+
+## Approach
+
+```mermaid
+flowchart TD
+  subgraph state["state — dict for O(1) lookup, doubly linked list for O(1) reorder"]
+    H["head — least recent"] --- N["... nodes ..."] --- T["tail — most recent"]
+  end
+  subgraph get["get(key)"]
+    G1{"key in cache?"} -- no --> G2(["return -1"])
+    G1 -- yes --> G3["_remove_node then _add_node — move it beside the tail"]
+    G3 --> G4(["return node.value"])
+  end
+  subgraph put["put(key, value)"]
+    P1{"key in cache?"} -- yes --> P2["update value, _remove_node then _add_node"]
+    P1 -- no --> P3["new Node, cache[key] = node, _add_node"]
+    P3 --> P4{"len(cache) > capacity?"}
+    P4 -- yes --> P5["evict head.next — the least recent — and del cache[lru.key]"]
+    P4 -- no --> P6(["done"])
+    P2 --> P6
+    P5 --> P6
+  end
+  state --> get
+  state --> put
+```
+
+## Solution
+
+```py
+class Node:
+    def __init__(self, key=0, value=0):
+        self.key = key
+        self.value = value
+        self.prev = None
+        self.next = None
+
+
+class LRUCache:
+    def __init__(self, capacity: int):
+        self.capacity = capacity
+        self.cache = {}
+        self.head = Node()
+        self.tail = Node()
+        self.head.next = self.tail  # pyright: ignore[reportAttributeAccessIssue]
+        self.tail.prev = self.head  # pyright: ignore[reportAttributeAccessIssue]
+
+    def _remove_node(self, node):
+        prev_node = node.prev
+        next_node = node.next
+        prev_node.next = next_node
+        next_node.prev = prev_node
+
+    def _add_node(self, node):
+        # Insert just before the tail (most recent at end)
+        node.next = self.tail
+        node.prev = self.tail.prev
+        self.tail.prev.next = node  # pyright: ignore[reportAttributeAccessIssue]
+        self.tail.prev = node
+
+    def get(self, key: int) -> int:
+        if key not in self.cache:
+            return -1
+
+        node = self.cache[key]
+        self._remove_node(node)
+        self._add_node(node)
+        return node.value
+
+    def put(self, key: int, value: int) -> None:
+        if key in self.cache:
+            node = self.cache[key]
+            node.value = value
+            self._remove_node(node)
+            self._add_node(node)
+        else:
+            new_node = Node(key, value)
+            self.cache[key] = new_node
+            self._add_node(new_node)
+
+            if len(self.cache) > self.capacity:
+                lru = self.head.next  # Oldest node is next to the head
+                self._remove_node(lru)
+                del self.cache[lru.key]  # pyright: ignore[reportAttributeAccessIssue]
+
+
+# Your LRUCache object will be instantiated and called as such:
+# obj = LRUCache(capacity)
+# param_1 = obj.get(key)
+# obj.put(key,value)
+```
+
+## Explanation
+
+<YouTube id="7ABFKPK2hD4" title="LRU Cache - Twitch Interview Question - Leetcode 146" />

apps/docs/content/(hash-table)/1832-check-if-the-sentence-is-pangram.mdx

@@ -22,6 +22,19 @@ sidebar:
 - `1 <= sentence.length <= 1000`
 - `sentence` consists of lowercase English letters.
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["checkIfPangram(sentence)"]) --> I["freq = {}"]
+  I --> L{"more char c?"}
+  L -- yes --> C["freq[c] = (freq[c] or 0) + 1"]
+  C --> L
+  L -- no --> Q{"distinct keys in freq == 26?"}
+  Q -- yes --> R(["return true"])
+  Q -- no --> F(["return false"])
+```
+
 ## Solution
 
 ```js

apps/docs/content/(hash-table)/242-valid-anagram.mdx

@@ -25,6 +25,19 @@ What if the inputs contain Unicode characters? How would you adapt your solution
 - `1 <= s.length, t.length <= 5 * 10^4`
 - `s` and `t` consist of lowercase English letters.
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["isAnagram(s, t)"]) --> I["freq: count every char of s"]
+  I --> L{"more char c in t?"}
+  L -- no --> E(["return true"])
+  L -- yes --> Q{"freq[c] missing or 0?"}
+  Q -- yes --> R(["return false — t has a char s cannot cover"])
+  Q -- no --> D["freq[c] -= 1"]
+  D --> L
+```
+
 ## Solution
 
 ```js
@@ -46,3 +59,7 @@ var isAnagram = function(s, t) {
   return true;
 };
 ```
+
+## Explanation
+
+<YouTube id="9UtInBqnCgA" title="Valid Anagram - Leetcode 242 - Python" />

apps/docs/content/(hash-table)/3-longest-substring-without-repeating-characters.mdx

@@ -0,0 +1,66 @@
+---
+title: '3. Longest Substring Without Repeating Characters'
+description: Given a string s, find the length of the longest substring without duplicate characters
+sidebar:
+  label: 'Longest Substring Without Repeating Characters'
+  badge: 'Medium'
+---
+
+<Badge variant="accent">Sliding Window</Badge>
+
+### Example 1:
+- Input: `s = "abcabcbb"`
+- Output: `3`
+- Explanation: The answer is "abc", with the length of `3`. Note that "bca" and "cab" are also correct answers.
+
+### Example 2:
+- Input: `s = "bbbbb"`
+- Output: `1`
+- Explanation: The answer is "b", with the length of `1`.
+
+### Example 3:
+- Input: `s = "pwwkew"`
+- Output: `3`
+- Explanation: The answer is "wke", with the length of `3`. Notice that the answer must be a substring, "pwke" is a subsequence and not a substring.
+
+### Constraints:
+
+- `0 <= s.length <= 10^5`
+- `s` consists of English letters, digits, symbols and spaces.
+
+## Approach
+
+```mermaid
+flowchart TD
+  S(["lengthOfLongestSubstring(s)"]) --> I["left = 0, ans = 0, window = set()"]
+  I --> L{"more (right, c) in s?"}
+  L -- no --> E(["return ans"])
+  L -- yes --> W{"c already in window?"}
+  W -- yes --> P["window.remove(s[left]); left += 1 — shrink until c is free"]
+  P --> W
+  W -- no --> A["window.add(c)"]
+  A --> M["ans = max(ans, right - left + 1)"]
+  M --> L
+```
+
+## Solution
+
+```py
+class Solution:
+    def lengthOfLongestSubstring(self, s: str) -> int:
+        left = 0
+        ans = 0
+        window = set()
+        for right, c in enumerate(s):
+            while c in window:
+                window.remove(s[left])
+                left += 1
+            window.add(c)
+            ans = max(ans, right - left + 1)
+
+        return ans
+```
+
+## Explanation
+
+<YouTube id="wiGpQwVHdE0" title="Longest Substring Without Repeating Characters - Leetcode 3 - Python" />

apps/docs/content/(hash-table)/387-first-unique-character-in-a-string.mdx

@@ -26,6 +26,18 @@ sidebar:
 - `1 <= s.length <= 10^5`
 - `s` consists of only lowercase English letters.
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["firstUniqChar(s)"]) --> F["Pass 1 — freq: count every char of s"]
+  F --> L{"Pass 2 — more i in 0..len-1?"}
+  L -- no --> E(["return -1"])
+  L -- yes --> Q{"freq[s[i]] == 1?"}
+  Q -- yes --> R(["return i — leftmost wins because the scan is in order"])
+  Q -- no --> L
+```
+
 ## Solution
 
 ```js
@@ -45,3 +57,7 @@ var firstUniqChar = function (s) {
   return -1;
 };
 ```
+
+## Explanation
+
+<YouTube id="rBENYgWy3xU" title="First Unique Character in a String - Leetcode 387 - Python" />

apps/docs/content/(hash-table)/424-longest-repeating-character-replacement.mdx

@@ -0,0 +1,66 @@
+---
+title: '424. Longest Repeating Character Replacement'
+description: You are given a string s and an integer k. You can choose any character of the string and change it to any other uppercase English character. You can perform this operation at most k times
+sidebar:
+  label: 'Longest Repeating Character Replacement'
+  badge: 'Medium'
+---
+
+<Badge variant="accent">Sliding Window</Badge>
+
+### Example 1:
+- Input: `s = "ABAB", k = 2`
+- Output: `4`
+- Explanation: Replace the two 'A's with two 'B's or vice versa.
+
+### Example 2:
+- Input: `s = "AABABBA", k = 1`
+- Output: `4`
+- Explanation: Replace the one 'A' in the middle with 'B' and form "AABBBBA". The substring "BBBB" has the longest repeating letters, which is `4`. There may exists other ways to achieve this answer too.
+
+### Constraints:
+
+- `1 <= s.length <= 10^5`
+- `s` consists of only uppercase English letters.
+- `0 <= k <= s.length`
+
+## Approach
+
+```mermaid
+flowchart TD
+  S(["characterReplacement(s, k)"]) --> I["count = {}, res = 0, l = 0, maxF = 0"]
+  I --> L{"more (r, c) in s?"}
+  L -- no --> E(["return res"])
+  L -- yes --> A["count[c] += 1; maxF = max(maxF, count[c])"]
+  A --> W{"(r - l + 1) - maxF > k? — too many chars to replace"}
+  W -- yes --> P["count[s[l]] -= 1; l += 1 — shrink from the left"]
+  P --> W
+  W -- no --> M["res = max(res, r - l + 1)"]
+  M --> L
+```
+
+## Solution
+
+```py
+class Solution:
+    def characterReplacement(self, s: str, k: int) -> int:
+        count = {}
+        res = 0
+
+        l = 0
+        maxF = 0
+        for r, c in enumerate(s):
+            count[c] = 1 + count.get(c, 0)
+            maxF = max(maxF, count[c])
+
+            while (r - l + 1) - maxF > k:
+                count[s[l]] -= 1
+                l += 1
+
+            res = max(res, r - l + 1)
+        return res
+```
+
+## Explanation
+
+<YouTube id="gqXU1UyA8pk" title="Longest Repeating Character Replacement - Leetcode 424 - Python" />

apps/docs/content/(hash-table)/451-sort-characters-by-frequency.mdx

@@ -28,6 +28,18 @@ sidebar:
 - `1 <= s.length <= 5 * 10^5`
 - `s` consists of uppercase and lowercase English letters and digits.
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["frequencySort(s)"]) --> F["freq: count every char of s"]
+  F --> C["split s and sort with a two-key comparator"]
+  C --> K1["primary: freq[b] - freq[a] — most frequent first"]
+  C --> K2["tie-break: a.localeCompare(b) — alphabetical"]
+  K1 --> R(["join and return"])
+  K2 --> R
+```
+
 ## Solution
 
 ```js
@@ -47,3 +59,7 @@ var frequencySort = function (s) {
     .join("");
 };
 ```
+
+## Explanation
+
+<YouTube id="OXdXc9HTrIg" title="Sort Characters By Frequency - Leetcode 451 - Python" />

apps/docs/content/(hash-table)/567-permutation-in-string.mdx

@@ -0,0 +1,73 @@
+---
+title: '567. Permutation in String'
+description: Given two strings s1 and s2, return true if s2 contains a permutation of s1, or false otherwise
+sidebar:
+  label: 'Permutation in String'
+  badge: 'Medium'
+---
+
+<Badge variant="accent">Sliding Window</Badge>
+
+### Example 1:
+- Input: `s1 = "ab", s2 = "eidbaooo"`
+- Output: `true`
+- Explanation: `s2` contains one permutation of `s1` ("ba").
+
+### Example 2:
+- Input: `s1 = "ab", s2 = "eidboaoo"`
+- Output: `false`
+
+### Constraints:
+
+- `1 <= s1.length, s2.length <= 10^4`
+- `s1` and `s2` consist of lowercase English letters.
+
+## Approach
+
+```mermaid
+flowchart TD
+  S(["checkInclusion(s1, s2)"]) --> G{"k = len(s1) > len(s2)?"}
+  G -- yes --> X(["return False"])
+  G -- no --> I["need = Counter(s1), window = Counter()"]
+  I --> L{"more (right, c) in s2?"}
+  L -- no --> E(["return False"])
+  L -- yes --> A["window[c] += 1"]
+  A --> B{"right >= k? — window is now longer than k"}
+  B -- yes --> C["drop s2[right - k] from window, deleting the key at 0"]
+  B -- no --> Q
+  C --> Q{"window == need?"}
+  Q -- yes --> R(["return True"])
+  Q -- no --> L
+```
+
+## Solution
+
+```py
+from collections import Counter
+
+
+class Solution:
+    def checkInclusion(self, s1: str, s2: str) -> bool:
+        k = len(s1)
+        if k > len(s2):
+            return False
+
+        need = Counter(s1)
+        window = Counter()
+
+        for right, c in enumerate(s2):
+            window[c] += 1
+            if right >= k:
+                left_char = s2[right - k]
+                window[left_char] -= 1
+                if window[left_char] == 0:
+                    del window[left_char]
+            if window == need:
+                return True
+
+        return False
+```
+
+## Explanation
+
+<YouTube id="UbyhOgBN834" title="Permutation in String - Leetcode 567 - Python" />

apps/docs/content/(linked-list)/2-add-two-numbers.mdx

@@ -0,0 +1,57 @@
+---
+title: '2. Add Two Numbers'
+description: You are given two non-empty linked lists representing two non-negative integers. The digits are stored in reverse order, and each of their nodes contains a single digit. Add the two numbers and return the sum as a linked list
+sidebar:
+  label: 'Add Two Numbers'
+  badge: 'Medium'
+---
+
+<Badge variant="accent">Hybrid Structures</Badge>
+
+### Example 1:
+- Input: `l1 = [2,4,3], l2 = [5,6,4]`
+- Output: `[7,0,8]`
+- Explanation: `342 + 465 = 807`.
+
+### Example 2:
+- Input: `l1 = [0], l2 = [0]`
+- Output: `[0]`
+
+### Example 3:
+- Input: `l1 = [9,9,9,9,9,9,9], l2 = [9,9,9,9]`
+- Output: `[8,9,9,9,0,0,0,1]`
+
+### Constraints:
+
+- The number of nodes in each linked list is in the range [1, 100].
+- `0 <= Node.val <= 9`
+- It is guaranteed that the list represents a number that does not have leading zeros.
+
+## Solution
+
+```py
+# Definition for singly-linked list.
+# class ListNode:
+#     def __init__(self, val=0, next=None):
+#         self.val = val
+#         self.next = next
+class Solution:
+    def addTwoNumbers(
+        self, l1: Optional[ListNode], l2: Optional[ListNode]
+    ) -> Optional[ListNode]:
+
+        dummy = ListNode()
+        carry = 0
+        curr = dummy
+        while l1 or l2 or carry:
+            s = (l1.val if l1 else 0) + (l2.val if l2 else 0) + carry
+            carry, val = divmod(s, 10)
+
+            curr.next = ListNode(val)
+            curr = curr.next
+
+            l1 = l1.next if l1 else None
+            l2 = l2.next if l2 else None
+
+        return dummy.next
+```

apps/docs/content/(linked-list)/206-reverse-linked-list.mdx

@@ -0,0 +1,67 @@
+---
+title: '206. Reverse Linked List'
+description: Given the head of a singly linked list, reverse the list, and return the reversed list
+sidebar:
+  label: 'Reverse Linked List'
+  badge: 'Easy'
+---
+
+<Badge variant="accent">Linked List</Badge>
+
+::::warning
+A linked list can be reversed either iteratively or recursively. Could you implement both?
+::::
+
+### Example 1:
+- Input: `head = [1,2,3,4,5]`
+- Output: `[5,4,3,2,1]`
+
+### Example 2:
+- Input: `head = [1,2]`
+- Output: `[2,1]`
+
+### Example 3:
+- Input: `head = []`
+- Output: `[]`
+
+### Constraints:
+
+- The number of nodes in the list is the range [0, 5000].
+- `-5000 <= Node.val <= 5000`
+
+## Approach
+
+```mermaid
+flowchart TD
+  S(["reverseList(head)"]) --> I["prev = None, curr = head"]
+  I --> W{"curr?"}
+  W -- no --> E(["return prev — the old tail is the new head"])
+  W -- yes --> A["next_ = curr.next — save it before overwriting"]
+  A --> B["curr.next = prev — flip the link"]
+  B --> C["prev = curr; curr = next_ — step forward"]
+  C --> W
+```
+
+## Solution
+
+```py
+# Definition for singly-linked list.
+# class ListNode:
+#     def __init__(self, val=0, next=None):
+#         self.val = val
+#         self.next = next
+class Solution:
+    def reverseList(self, head: Optional[ListNode]) -> Optional[ListNode]:
+        prev = None
+        curr = head
+        while curr:
+            next_ = curr.next
+            curr.next = prev
+            prev = curr
+            curr = next_
+        return prev
+```
+
+## Explanation
+
+<YouTube id="G0_I-ZF0S38" title="Reverse Linked List - Iterative AND Recursive - Leetcode 206 - Python" />

apps/docs/content/(linked-list)/21-merge-two-sorted-lists.mdx

@@ -0,0 +1,87 @@
+---
+title: '21. Merge Two Sorted Lists'
+description: You are given the heads of two sorted linked lists list1 and list2
+sidebar:
+  label: 'Merge Two Sorted Lists'
+  badge: 'Easy'
+---
+
+<Badge variant="accent">Linked List</Badge>
+
+### Example 1:
+- Input: `list1 = [1,2,4], list2 = [1,3,4]`
+- Output: `[1,1,2,3,4,4]`
+
+### Example 2:
+- Input: `list1 = [], list2 = []`
+- Output: `[]`
+
+### Example 3:
+- Input: `list1 = [], list2 = [0]`
+- Output: `[0]`
+
+### Constraints:
+
+- The number of nodes in both lists is in the range [0, 50].
+- `-100 <= Node.val <= 100`
+- Both list1 and list2 are sorted in non-decreasing order.
+
+## Approach
+
+```mermaid
+flowchart TD
+  S(["mergeTwoLists(list1, list2)"]) --> G1{"list1 empty?"}
+  G1 -- yes --> X1(["return list2"])
+  G1 -- no --> G2{"list2 empty?"}
+  G2 -- yes --> X2(["return list1"])
+  G2 -- no --> H["head = the smaller first node; advance that list"]
+  H --> C["current = head"]
+  C --> W{"list1 and list2 both non-empty?"}
+  W -- yes --> P["current.next = the smaller node; advance that list; current = current.next"]
+  P --> W
+  W -- no --> T["current.next = list1 or list2 — append the leftover tail"]
+  T --> E(["return head"])
+```
+
+## Solution
+
+```py
+# Definition for singly-linked list.
+# class ListNode:
+#     def __init__(self, val=0, next=None):
+#         self.val = val
+#         self.next = next
+class Solution:
+    def mergeTwoLists(
+        self, list1: Optional[ListNode], list2: Optional[ListNode]
+    ) -> Optional[ListNode]:
+
+        if not list1:
+            return list2
+        if not list2:
+            return list1
+
+        if list1.val < list2.val:
+            head = list1
+            list1 = list1.next
+        else:
+            head = list2
+            list2 = list2.next
+
+        current = head
+        while list1 and list2:
+            if list1.val < list2.val:
+                current.next = list1
+                list1 = list1.next
+            else:
+                current.next = list2
+                list2 = list2.next
+            current = current.next
+
+        current.next = list1 or list2
+        return head
+```
+
+## Explanation
+
+<YouTube id="XIdigk956u0" title="Merge Two Sorted Lists - Leetcode 21 - Python" />

apps/docs/content/(stack)/155-min-stack.mdx

@@ -0,0 +1,82 @@
+---
+title: '155. Min Stack'
+description: Design a stack that supports push, pop, top, and retrieving the minimum element in constant time
+sidebar:
+  label: 'Min Stack'
+  badge: 'Medium'
+---
+
+<Badge variant="accent">Stack</Badge>
+
+::::warning
+You must implement a solution with O(1) time complexity for each function.
+::::
+
+### Example 1:
+- Input: ``
+- Output: ``
+- Explanation: Input `["MinStack","push","push","push","getMin","pop","top","getMin"]` [[],[-2],[0],[-3],[],[],[],[]] Output `[null,null,null,null,-3,null,0,-2]` Explanation MinStack minStack = new MinStack(); minStack.push(-2); minStack.push(0); minStack.push(-3); minStack.getMin(); // return `-3` minStack.pop(); minStack.top();    // return `0` minStack.getMin(); // return `-2`
+
+### Constraints:
+
+- `-2^31 <= val <= 2^31 - 1`
+- Methods pop, top and getMin operations will always be called on non-empty stacks.
+- At most 3 * 10^4 calls will be made to push, pop, top, and getMin.
+
+## Approach
+
+```mermaid
+flowchart TD
+  subgraph state["two parallel stacks — same depth, always"]
+    A["stack — the values"]
+    B["minStack — the minimum as of that depth"]
+  end
+  state --> P
+  subgraph P["push(value)"]
+    P1["stack.append(value)"] --> P2["minStack.append(min(value, minStack[-1]))"]
+  end
+  state --> O
+  subgraph O["pop()"]
+    O1["pop both stacks together"]
+  end
+  state --> R
+  subgraph R["top() / getMin() — O(1)"]
+    R1(["stack[-1] / minStack[-1]"])
+  end
+```
+
+## Solution
+
+```py
+class MinStack:
+    def __init__(self):
+        self.stack = []
+        self.minStack = []
+
+    def push(self, value: int) -> None:
+        self.stack.append(value)
+        value = min(value, self.minStack[-1] if self.minStack else value)
+        self.minStack.append(value)
+
+    def pop(self) -> None:
+        self.stack.pop()
+        self.minStack.pop()
+
+    def top(self) -> int:
+        return self.stack[-1]
+
+    def getMin(self) -> int:
+        return self.minStack[-1]
+
+
+# Your MinStack object will be instantiated and called as such:
+# obj = MinStack()
+# obj.push(value)
+# obj.pop()
+# param_3 = obj.top()
+# param_4 = obj.getMin()
+```
+
+## Explanation
+
+<YouTube id="qkLl7nAwDPo" title="Design Min Stack - Amazon Interview Question - Leetcode 155 - Python" />

apps/docs/content/(string)/20-valid-parentheses.mdx

@@ -0,0 +1,76 @@
+---
+title: '20. Valid Parentheses'
+description: Given a string s containing just the characters '(', ')', '{', '}', '[' and ']', determine if the input string is valid
+sidebar:
+  label: 'Valid Parentheses'
+  badge: 'Easy'
+---
+
+<Badge variant="accent">Stack</Badge>
+
+### Example 1:
+- Input: `s = "()"`
+- Output: `true`
+
+### Example 2:
+- Input: `s = "()[]{}"`
+- Output: `true`
+
+### Example 3:
+- Input: `s = "(]"`
+- Output: `false`
+
+### Example 4:
+- Input: `s = "([])"`
+- Output: `true`
+
+### Example 5:
+- Input: `s = "([)]"`
+- Output: `false`
+
+### Constraints:
+
+- `1 <= s.length <= 10^4`
+- `s` consists of parentheses only '()[]{}'.
+
+## Approach
+
+```mermaid
+flowchart TD
+  S(["isValid(s)"]) --> I["stack = [], pairs = closer to opener"]
+  I --> L{"more char c?"}
+  L -- no --> Z{"stack empty?"}
+  Z -- yes --> T(["return True"])
+  Z -- no --> U(["return False — unclosed openers left"])
+  L -- yes --> Q{"c is a closer?"}
+  Q -- no --> A["stack.append(c) — it is an opener"]
+  A --> L
+  Q -- yes --> M{"stack non-empty and pairs[c] == stack[-1]?"}
+  M -- yes --> P["stack.pop() — matched"]
+  P --> L
+  M -- no --> F(["return False — mismatch"])
+```
+
+## Solution
+
+```py
+class Solution:
+    def isValid(self, s: str) -> bool:
+        stack = []
+        pairs = {")": "(", "]": "[", "}": "{"}
+
+        for c in s:
+            if c in pairs:
+                if stack and pairs[c] == stack[-1]:
+                    stack.pop()
+                else:
+                    return False
+            else:
+                stack.append(c)
+
+        return True if not stack else False
+```
+
+## Explanation
+
+<YouTube id="WTzjTskDFMg" title="Valid Parentheses - Stack - Leetcode 20 - Python" />

apps/docs/content/(two-pointers)/125-valid-palindrome.mdx

@@ -28,6 +28,20 @@ sidebar:
 - `1 <= s.length <= 2 * 10^5`
 - `s` consists only of printable ASCII characters.
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["isPalindrome(s)"]) --> N["normal = s stripped of non-alphanumerics, lowercased"]
+  N --> I["i = 0, j = normal.length - 1"]
+  I --> W{"i < j?"}
+  W -- no --> E(["return true — the pointers met in the middle"])
+  W -- yes --> Q{"normal[i] != normal[j]?"}
+  Q -- yes --> F(["return false"])
+  Q -- no --> A["i++; j-- — step both inward"]
+  A --> W
+```
+
 ## Solution
 
 ```js
@@ -49,3 +63,7 @@ var isPalindrome = function(s) {
   return true;
 };
 ```
+
+## Explanation
+
+<YouTube id="jJXJ16kPFWg" title="Valid Palindrome - Leetcode 125 - Python" />

apps/docs/content/(two-pointers)/344-reverse-string.mdx

@@ -25,6 +25,18 @@ You must do this by modifying the input array in-place with O(1) extra memory.
 - `1 <= s.length <= 10^5`
 - `s[i]` is a printable ascii character.
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["reverseString(s)"]) --> I["i = 0, j = s.length - 1"]
+  I --> W{"i < j?"}
+  W -- no --> E(["done — s was reversed in place, nothing returned"])
+  W -- yes --> A["swap s[i] and s[j]"]
+  A --> B["i++; j--"]
+  B --> W
+```
+
 ## Solution
 
 ```js
@@ -43,3 +55,7 @@ var reverseString = function(s) {
     }   
 };
 ```
+
+## Explanation
+
+<YouTube id="_d0T_2Lk2qA" title="Reverse String - 3 Ways - Leetcode 344 - Python" />

apps/docs/content/(two-pointers)/392-is-subsequence.mdx

@@ -26,6 +26,22 @@ Suppose there are lots of incoming s, say s1, s2, ..., sk where k >= 10^9, and y
 - `0 <= t.length <= 10^4`
 - `s` and `t` consist only of lowercase English letters.
 
+## Approach
+
+```mermaid
+flowchart TD
+  S(["isSubsequence(s, t)"]) --> G{"len(s) > len(t)?"}
+  G -- yes --> X(["return False"])
+  G -- no --> I["i = 0 into s, j = 0 into t"]
+  I --> W{"i < len(s) and j < len(t)?"}
+  W -- no --> E(["return i == len(s) — every char of s was matched"])
+  W -- yes --> Q{"s[i] == t[j]?"}
+  Q -- yes --> A["i += 1 — consume the match"]
+  Q -- no --> B["j += 1 — always advance t"]
+  A --> B
+  B --> W
+```
+
 ## Solution
 
 ```py
@@ -42,3 +58,7 @@ class Solution:
 
         return i == len(s)
 ```
+
+## Explanation
+
+<YouTube id="99RVfqklbCE" title="Is Subsequence - Leetcode 392" />