leetcode/apps/docs/content/todo.md

hidden

hidden
true

todo — draining the review backlog

every solution in work/1 owes a blind re-solve at work/3 and a second one at work/7. both buckets are empty, so 36 problems owe 72 re-solves and not one has been started. snapshot: reviewqueues("2026-08-31") from apps/cli/work.ts — what bun run pick shows — cross-checked against /api/stats.

at 10 problems a day including the campaign's own load, 72 re-solves do not fit in one week: this week absorbs 38, the +3 rung finishes sat sep 5 (campaign day 20), the last +7 clears fri sep 18 (day 33), and normal srs resumes sat sep 19 — see when standard srs resumes. saturday gates are deliberately left out of the count.

rules this plan follows

• ≤ 10 problems a day, everything counted. each day's new core set and the on-time reviews it earns are booked first; the backlog takes what is left — 7 slots on a light day, 2 on a heavy one.
• coverage before depth. all 36 +3 re-solves come first — one blind pass over everything — before any problem gets its +7.
• oldest first. lc 387 and lc 1365 have been sitting 50 days.
• ≥ 3 days between a problem's own +3 and +7. two re-solves in one day test nothing.
• sunday is not a slot. workingday() never mints one; neither does this list.
• re-solve blind, then push. bun run pick (tab cycles to the 3-day and 7-day sections) scaffolds straight into work/3/work/7. do not open work/1 first. pushing the file is the log: close-solved.yml posts the bucket to /admin/solved and d1 moves the problem up the ladder.

this week

mon aug 31 — 3 new + 7 backlog = 10/10

topic 08 — linked list (#10, closed today)

bucket	problem	difficulty	where	why
work/1	lc 21 · merge two sorted lists	easy	linked list	done today
work/1	lc 206 · reverse linked list	easy	linked list	done today
work/1	lc 141 · linked list cycle	easy	hash table	done today
work/3	lc 387 · first unique character in a string	easy	hash table	50d stale — first solved 2026-07-12
work/3	lc 1365 · how many numbers are smaller than the current number	easy	array	50d stale — first solved 2026-07-12
work/3	lc 268 · missing number	easy	array	14d stale — first solved 2026-08-17
work/3	lc 344 · reverse string	easy	two pointers	14d stale — first solved 2026-08-17
work/3	lc 1426 · counting elements	easy	array	14d stale — first solved 2026-08-17
work/3	lc 1832 · check if the sentence is pangram	easy	hash table	14d stale — first solved 2026-08-17
work/3	lc 15 · 3sum	medium	array	13d stale — first solved 2026-08-18

tue sep 1 — 2 new + 8 backlog = 10/10

topic 09 — hybrid structures (#11)

bucket	problem	difficulty	where	why
work/1	lc 146 · lru cache	medium	topic 09 — hybrid structures	new core problem
work/1	lc 2 · add two numbers	medium	topic 09 — hybrid structures	new core problem
work/3	lc 49 · group anagrams	medium	array	13d stale — first solved 2026-08-18
work/3	lc 125 · valid palindrome	easy	two pointers	13d stale — first solved 2026-08-18
work/3	lc 167 · two sum ii - input array is sorted	medium	array	13d stale — first solved 2026-08-18
work/3	lc 217 · contains duplicate	easy	array	13d stale — first solved 2026-08-18
work/3	lc 238 · product of array except self	medium	array	13d stale — first solved 2026-08-18
work/3	lc 242 · valid anagram	easy	hash table	13d stale — first solved 2026-08-18
work/3	lc 303 · range sum query - immutable	easy	array	13d stale — first solved 2026-08-18
work/3	lc 560 · subarray sum equals k	medium	array	13d stale — first solved 2026-08-18

wed sep 2 — 3 new + 7 backlog = 10/10

topic 10 — matrix index math (#12)

bucket	problem	difficulty	where	why
work/1	lc 48 · rotate image	medium	topic 10 — matrix index math	new core problem
work/1	lc 54 · spiral matrix	medium	topic 10 — matrix index math	new core problem
work/1	lc 73 · set matrix zeroes	medium	topic 10 — matrix index math	new core problem
work/3	lc 1413 · minimum value to get positive step by step sum	easy	array	13d stale — first solved 2026-08-18
work/3	lc 1480 · running sum of 1d array	easy	array	13d stale — first solved 2026-08-18
work/3	lc 33 · search in rotated sorted array	medium	array	12d stale — first solved 2026-08-19
work/3	lc 704 · binary search	easy	array	12d stale — first solved 2026-08-19
work/3	lc 875 · koko eating bananas	medium	array	12d stale — first solved 2026-08-19
work/3	lc 977 · squares of a sorted array	easy	array	12d stale — first solved 2026-08-19
work/3	lc 121 · best time to buy and sell stock	easy	array	11d stale — first solved 2026-08-20

thu sep 3 — 3 new + 3 on-time reviews + 4 backlog = 10/10

topic 11 — tree dfs (#13)

bucket	problem	difficulty	where	why
work/1	lc 226 · invert binary tree	easy	topic 11 — tree dfs	new core problem
work/1	lc 104 · maximum depth of binary tree	easy	topic 11 — tree dfs	new core problem
work/1	lc 236 · lowest common ancestor of a binary tree	medium	topic 11 — tree dfs	new core problem
work/3	lc 21 · merge two sorted lists	easy	linked list	on time — solved mon aug 31
work/3	lc 206 · reverse linked list	easy	linked list	on time — solved mon aug 31
work/3	lc 141 · linked list cycle	easy	hash table	on time — solved mon aug 31
work/3	lc 1 · two sum	easy	array	7d stale — first solved 2026-08-24
work/3	lc 11 · container with most water	medium	array	7d stale — first solved 2026-08-24
work/3	lc 392 · is subsequence	easy	two pointers	7d stale — first solved 2026-08-24
work/3	lc 3 · longest substring without repeating characters	medium	hash table	6d stale — first solved 2026-08-25

fri sep 4 — 3 new + 2 on-time reviews + 5 backlog = 10/10

topic 12 — tree bfs (#14)

bucket	problem	difficulty	where	why
work/1	lc 102 · binary tree level order traversal	medium	topic 12 — tree bfs	new core problem
work/1	lc 199 · binary tree right side view	medium	topic 12 — tree bfs	new core problem
work/1	lc 1448 · count good nodes in binary tree	medium	topic 12 — tree bfs	new core problem
work/3	lc 146 · lru cache	medium	topic 09 — hybrid structures	on time — solved tue sep 1
work/3	lc 2 · add two numbers	medium	topic 09 — hybrid structures	on time — solved tue sep 1
work/3	lc 424 · longest repeating character replacement	medium	hash table	6d stale — first solved 2026-08-25
work/3	lc 567 · permutation in string	medium	hash table	6d stale — first solved 2026-08-25
work/3	lc 35 · search insert position	easy	array	5d stale — first solved 2026-08-26
work/3	lc 74 · search a 2d matrix	medium	array	5d stale — first solved 2026-08-26
work/3	lc 153 · find minimum in rotated sorted array	medium	array	5d stale — first solved 2026-08-26

sat sep 5 — 3 on-time reviews + 7 backlog = 10/10

no new topic — saturday.

bucket	problem	difficulty	where	why
work/3	lc 48 · rotate image	medium	topic 10 — matrix index math	on time — solved wed sep 2
work/3	lc 54 · spiral matrix	medium	topic 10 — matrix index math	on time — solved wed sep 2
work/3	lc 73 · set matrix zeroes	medium	topic 10 — matrix index math	on time — solved wed sep 2
work/3	lc 20 · valid parentheses	easy	string	4d stale — first solved 2026-08-27
work/3	lc 150 · evaluate reverse polish notation	medium	array	4d stale — first solved 2026-08-27
work/3	lc 155 · min stack	medium	stack	4d stale — first solved 2026-08-27
work/3	lc 739 · daily temperatures	medium	array	3d stale — first solved 2026-08-28
work/3	lc 853 · car fleet	medium	array	3d stale — first solved 2026-08-28
work/7	lc 387 · first unique character in a string	easy	hash table	+3 was mon aug 31
work/7	lc 1365 · how many numbers are smaller than the current number	easy	array	+3 was mon aug 31

sun sep 6 — rest day

nothing. by design.

carry-over

the 34 re-solves that do not fit this week, at the same cap and the same rules. booked = that day's new core set plus the on-time reviews it owes.

day	topic	booked	backlog re-solves	load
mon sep 7	topic 13 — binary search tree	10	none — the day is already full	10/10
tue sep 8	topic 14 — heap / priority queue	7	work/7 — lc 268, lc 344, lc 1426	10/10
wed sep 9	topic 15 — trie	5	work/7 — lc 1832, lc 15, lc 49, lc 125, lc 167	10/10
thu sep 10	topic 16 — graph dfs	9	work/7 — lc 217	10/10
fri sep 11	topic 17 — graph bfs	8	work/7 — lc 238, lc 242	10/10
sat sep 12	none — saturday	2	work/7 — lc 303, lc 560, lc 1413, lc 1480, lc 33, lc 704, lc 875, lc 977	10/10
mon sep 14	topic 18 — topological sort	10	none — the day is already full	10/10
tue sep 15	topic 19 — union-find	5	work/7 — lc 121, lc 1, lc 11, lc 392, lc 3	10/10
wed sep 16	topic 20 — backtracking	5	work/7 — lc 424, lc 567, lc 35, lc 74, lc 153	10/10
thu sep 17	topic 21 — 1-d dynamic programming	8	work/7 — lc 20, lc 150	10/10
fri sep 18	topic 22 — multi-d / grid dp	7	work/7 — lc 155, lc 739, lc 853	10/10

when standard srs resumes

• sat sep 5 (campaign day 20) — the +3 rung is empty. every one of the 36 has been recalled blind once; nothing is unretrieved any more.
• fri sep 18 (campaign day 33) — the last backlog +7 lands. work/3 and work/7 hold a file for all 36, so reviewqueues() owes nothing; the 28 that have a d1 row are retired on the ladder, and the 8 listed at the bottom of this file never had one.
• sat sep 19 onward — standard srs. the only reviews left are the ones the campaign itself minted: solve a problem, its +3 and +7 come back on workingday(solve + 3) and workingday(solve + 7), the digest asks for ≤ review_cap (3) a day, levelreviews() pushes any excess forward, and bun run pick's 3-day and 7-day sections show only what is due. no dated plan needed after this — the digest is the plan.

campaign-native load per working day once the backlog is gone, all of it inside the cap:

day	topic	new	work/3	work/7	load
sat sep 19	review-only	0	3	0	3/10
mon sep 21	topic 23 — greedy	3	5	2	10/10
tue sep 22	topic 24 — intervals	3	1	2	6/10
wed sep 23	review-only	0	0	3	3/10
thu sep 24	review-only	0	3	3	6/10
fri sep 25	review-only	0	3	3	6/10
sat sep 26	review-only	0	0	0	0/10
mon sep 28	review-only	0	0	3	3/10
tue sep 29	review-only	0	0	3	3/10

two things to watch there. mon sep 21 is 10/10 on campaign load alone — topic 23's core set, two +7s, and five of the six +3s that land there (thursday's three slide off sunday onto monday, friday's three land natively, and the sixth is pushed to tuesday exactly as levelreviews() would) — so one slipped day re-opens a backlog. and from wed sep 30 the ladder runs dry: apps/api/data/schedule.json books its last topic tue sep 22, so the review-only weeks that close the campaign are drills, optional sets and the weekly gates, none of which this file counts.

not reviews

three july scaffolds in work/1 are still empty, so they owe no re-solve — they are unsolved problems, and bun run pick still lists them in its new-problem section:

problem	difficulty
lc 347 · top k frequent elements	medium
lc 451 · sort characters by frequency	medium
lc 1636 · sort array by increasing frequency	easy

what this plan exposes

• 8 of the 36 have no problem issue (387, 1365, 268, 344, 1426, 1832, 1413, 1480) — pre-campaign solves. the catalog reconcile never invents rows, so d1 does not know them and no digest can ever surface them. bun run pick is the only thing that will ask for these reviews.
• /api/stats reports a stage that does not exist: ladder: { new: 128, "+3": 3, "+7": 0, retired: 0, "+2": 30 }. +2 is not in stages (apps/api/src/srs.ts), so 30 rows sit on a pre-0002 rung — which is why the digest queue plateaus at 32 due instead of tracking the tables above. that wants a migration, not a re-solve.