Commit Graph
10 Commits
Author SHA1 Message Date
Mads LorentzenandClaude Fable 5 261b57edfd docs: move Ko-fi button below intro paragraph (#33)
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-05 19:38:56 +02:00
Mads LorentzenandClaude Fable 5 a7f20e101c docs: add Ko-fi Buy me a coffee button to top of README (#32)
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-05 19:38:11 +02:00
Mads LorentzenandClaude Fable 5 f9c1978c65 docs: add linkedin-search to install lists in README and SETUP (#28)
Follow-up to #20: the linkedin-search CLI landed without being added to
the bun install instructions. Included in both lists with a note that
the install is optional (zero runtime dependencies, dev types only).
Also updated the README file tree and Bun prerequisite wording to
reflect that the toolset is no longer Denmark-only.

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-05 10:59:08 +02:00
Mads LorentzenandClaude Fable 5 f15b9fb65d Tighten pre-approved permissions and untrack settings.local.json (#27)
* fix: move scoped permissions to settings.json, drop curl, untrack settings.local.json

Addresses #23:
- Remove pre-approved Bash(curl:*) - no agent workflow uses curl, and a
  toolkit that routinely feeds untrusted job postings to the model should
  not ship a pre-approved exfiltration-capable command
- Move shared permissions to .claude/settings.json (committed by
  convention) and scope them tighter: Bash(bun run:*) for the job portal
  CLIs, Bash(python/python3 salary_lookup.py:*) for salary lookups
- Untrack .claude/settings.local.json - it was committed despite being
  listed in .gitignore; the file stays local for personal overrides

Reported-by: @josealfonsomora

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* docs(setup): warn existing cloners about stale settings.local.json

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-05 10:56:17 +02:00
d8f38fe766 Add country-agnostic linkedin-search skill (#20)
A general-purpose, field-agnostic job-search skill built on LinkedIn's public
jobs-guest endpoints. Works for any market out of the box — location is an
explicit required flag (no country default). Zero runtime dependencies (bun only);
search + detail commands.

Includes a personal-use / Terms-of-Service note (automated access is against
LinkedIn's ToS — keep volume low, non-commercial).

(Pairs with the .gitignore fix in #21, which lets skills under .agents/ be tracked.)

Co-authored-by: Akhil Tripathi <kodabear@Akhils-MacBook-Pro.local>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
2026-06-29 20:26:19 +02:00
Mads LorentzenandClaude Opus 4.7 974b29ec3a refactor: fold /setup_docs into /setup as third onboarding path (#9)
* refactor: fold /setup_docs into /setup as third onboarding path

- Auto-detect documents/ at /setup Step 0; offer three paths (documents
  folder, single CV import, interview).
- Inline document-driven merge logic into setup.md as Path A: read-before-
  write, additive vs. conflicting bucket merge with per-conflict prompts,
  inference labeling for behavioral and style files, STAR-stub generation
  instead of fabrication.
- Step 3 substeps for skill files skip when Path A populated them; non-
  skill substeps (CLAUDE.md, cv/main_example.tex, search-queries.md)
  always run. Path C interview preserved verbatim. /setup --section <name>
  update flow preserved.
- Delete .claude/commands/setup_docs.md.
- Update /reset prompts to point only at /setup. Update documents/README.md
  to reference /setup throughout. Update README.md to drop the standalone
  /setup_docs section and add upskill/ to the skills tree.
- Remove the trailing comma left in .claude/settings.local.json after the
  PR #6 revert (was technically invalid JSON).

Single onboarding entry point. documents/ folder convention unchanged.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

* docs(README): finish readme alignment for the /setup fold

- Update onboarding-paths tip to reflect three paths (was "Either"
  before /setup_docs landed in PR #6).
- Add documents/ and upskill/ entries to the file tree, both relied
  on by the new /setup Path A and by /upskill respectively.
- Add a brief "Other commands" section briefly describing /expand and
  /upskill so users discover them without having to read individual
  command files. /reset stays in the existing "Starting over"
  subsection, now linked from the new section.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-29 09:57:33 +02:00
Bach 2ba441f3fb feat: add /setup_docs, /reset, /expand commands and /upskill skill
Adds four contributions from @Michael-Bach:

- /setup_docs - document-driven profile population from a documents/ folder (CV, LinkedIn export, diplomas, references, past applications). Idempotent merge with explicit additive vs. conflicting buckets and per-conflict prompts.
- /reset - typed-RESET confirmation gate for clearing profile data and/or documents folder contents.
- /expand - additive competency enrichment from documents and public URLs already in the profile (GitHub repos, portfolio sites), with web-searched syllabus lookups for named courses and certifications.
- /upskill - skill-gap analysis vs tracked jobs (or single URL), produces a prioritized heatmap and learning plan with year-tagged WebSearch queries.

Also adds the documents/ folder convention with README and gitignore entries for personal output files.

Closes #6
2026-04-29 09:30:08 +02:00
Mads LorentzenandClaude Opus 4.7 f62b9ec893 docs(README): soften "token-efficient" bullet to match reality
The original bullet ("Token-efficient. Reviewer receives drafts inline
rather than re-reading files. Verification runs once, at the end.") is
technically true for the reviewer-dispatch path but invites readers to
infer that the overall workflow uses fewer tokens than master. That is
not actually defensible: the new Step 5 (compile-and-inspect) spends
tokens on Read'ing both PDFs and on layout-fix iteration cycles, which
typically offsets the savings from inline draft passing and scoped
reviewer reads. End-to-end the workflow is roughly token-neutral, with
the variance dominated by how many layout fixes the CV needs.

Reframes the bullet as "Token-efficient reviewer dispatch" (which is
accurate and verifiable from the prompt structure) and adds an explicit
note that Step 5 spends some of those savings on PDF rendering. The
trade is real - tokens for fewer broken PDFs reaching the user - and
worth surfacing honestly rather than overclaiming.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-25 13:56:41 +02:00
Mads LorentzenandClaude Opus 4.7 00678f9669 docs: document PDF verification loop and relevance-weighted cutting
Updates README.md and SETUP.md to reflect the new /apply workflow
additions shipped in the previous commit:

- Bumps the /apply step list from 6 to 7, adding "Compile and inspect"
  between Revise and Present
- Adds a "What makes this workflow different" subsection highlighting the
  PDF verification loop, relevance-weighted CV cutting, drafter-reviewer
  separation, and token-efficient dispatching
- Updates prerequisites note to call out lualatex (CV) and xelatex (cover
  letter) explicitly, with the reason each engine is required
- Updates SETUP.md's LaTeX section to match (pdflatex -> lualatex for the
  CV, with the MiKTeX fontawesome5 caveat)

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-24 19:36:45 +02:00
AI Job SearchandClaude Opus 4.6 c66d599d75 Initial release: AI-powered job application framework
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-23 08:34:06 +01:00