career-os
CLAUDE.md0 starsChanged 55 days agoMITView on GitHub
This folder is a career agent. You (Claude Code) are the reasoning half; the Node scripts are the deterministic half. Read this file first in every session, then profile/career-context.md.
In this repository15 other files it gives its agents.
- applySkill
- career-pathSkill
- decideSkill
- huntSkill
- inboxSkill
- interview-prepSkill
- logSkill
- networkSkill
- offer-reviewSkill
- prefsSkill
- researchSkill
- studySkill
- tailorSkill
- trackSkill
- weeklySkill
The file
We split this CLAUDE.md into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 6
Never invent facts about the candidate: Every resume bullet, cover-letter claim and interview answer must trace to profile/.
≈ “Common” is an estimate: at least 1 in 5 of the 27,641 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Career OS — instructions for Claude Code
This folder is a career agent. You (Claude Code) are the reasoning half; the Node scripts are the deterministic half. Read this file first in every session, then profile/career-context.md.
Ground rules
- Never invent facts about the candidate. Every resume bullet, cover-letter claim and interview answer must trace to
profile/. Ask when something is missing. - Never submit an application, send a message, or accept terms on the candidate's behalf.
/applyprepares everything and drives the browser up to the final submit; the candidate clicks. Emails and LinkedIn messages are drafted, never sent. - Tailored files are derived. Edit
profile/master-resume.md, then regenerate with/tailor. One folder per job underjobs/<slug>/. - A plain fetch that says "position has been filled" is usually a decoy (JS-rendered ATS sites serve a static fallback). Verify in the Browser pane or via the adapter's
details()before calling a role closed. - Free/local tooling only by default. No paid APIs required.
- No em dashes (—) or en dashes (–) in anything the candidate will send. Commas, colons, full stops, "to" for ranges.
- Respect
profile/career-context.md → Confidentiality(for example: describe clients by sector, never by name).
Layout
profile/ master-resume.md · preferences.json (search prefs + priority_model) · certifications.md · story-bank.md · career-context.md · achievements.md · study/
sources/ one adapter per job source + config/ (seek, linkedin, phenom, oracle-orc, greenhouse, lever, ashby, gradconnection, mckinsey [browser], wagov [WA Government], apsjobs [browser], adzuna, indeed [browser])
lib/ schema.mjs (Job, normalizeJob, dedupe, closingDate) · http.mjs (polite fetch, cache) · store.mjs · score.mjs (prescore + upside/priority) · insights.mjs · md.mjs
hunt.mjs sources × queries → data/jobs.json + reports/hunt-<date>.md (--source, --query, --location, --new-only, --details N, --dry, --top)
tools/ render.mjs (resume.md → PDF/DOCX; --compact for one page; --merge for one-file applications) · dashboard.mjs · track.mjs · details.mjs · rescore.mjs · insights.mjs · shortlist.mjs · recheck.mjs · refresh.sh · setup.mjs · test.mjs
jobs/<slug>/ jd.md · brief.md · resume.md · cover-letter.md · interview-prep.md · research.md · network.md · notes.md · outputs/
data/ jobs.json · pipeline.json · cache/ · secrets.json (optional API keys)
reports/ hunt-<date>.md · shortlist-<date>.md · market-insights-<date>.md · career-path-<date>.md · dashboard.html
inbox/ drop JDs, transcripts, offer letters, question banks here for ingestion
.claude/skills/ the slash commands belowSlash commands
/hunt (fetch, rate fit, shortlist by priority) · /tailor <url|jobs/slug> (JD → brief + resume + cover letter → PDF/DOCX) · /interview-prep <slug> [mock] · /career-path · /network · /offer-review · /log <what you did> · /apply <slug> (stops before submit) · /track · /weekly · /prefs · /inbox (Gmail, draft only) · /research <company> · /decide <question> (deep career decision: live research, graded sources, kill criteria, decision log) · /study
Concepts
- Fit (Strong / Good / Stretch / Skip) = can the candidate get it. Rated by Claude from full descriptions. A "N+ years" line is a preference, not a wall; only 8+ years, principal/staff/manager titles, wrong domain or a true hard gate (citizenship, clearance) make a Skip.
- Priority (P1 / P2 / P3) = is it worth the effort. Computed by
lib/score.mjs upside()frompreferences.json → priority_model: brand direction vs current employer, interest alignment, pay, contract penalty. Never set by hand; edit the tiers instead. - One page for MBB and graduate programs (
--compact); two pages for industry roles. No profile paragraph on MBB/graduate CVs.
Running things
npm run setup # first run: creates profile/* from examples, runs offline tests
node hunt.mjs --new-only --details 15 # daily fetch
node tools/details.mjs --top 40 # full descriptions before rating
node tools/rescore.mjs && node tools/insights.mjs && node tools/shortlist.mjs && node tools/dashboard.mjs
node tools/render.mjs jobs/<slug> --compact # one-page PDF + DOCX
bash tools/refresh.sh # the whole chain (what a scheduler runs)
npm test # unit + live adapter smoke testsScheduling options are in README.md (local desktop task, launchd, cron, or a cloud routine over a private git repo).
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

