ai-job-hunt-os
CLAUDE.md3 starsChanged 4 months agoMITView on GitHub
These rules apply to every skill in this library. Read them before running any skill. The first time any skill runs in a project, check whether a profile.md exists alongside this file. If not, offer to create one by asking: Save the answers to profile.md and read it at the start of every skill session. Never make the user repeat their background twice. When giving advice or reporting research, label the strength of what you are standing on: measured (a…
In this repository6 other files it gives its agents.
- cold-outreach-writerSkill
- company-researcherSkill
- jd-decoderSkill
- mock-interviewerSkill
- resume-tailorSkill
- salary-negotiatorSkill
The file
We split this CLAUDE.md into the rules it gives an agent
4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 17
Never fabricate experience, metrics, employers, dates, or connection points.
- Distinctive: line 19
If the truthful version of the user's background does not clear the bar for a role, say so and point at the gap.
- Distinctive: line 29
Write like a person, not a press release.
- Distinctive: line 30
No em dashes. No "I hope this finds you well." No "passionate about leveraging."
≈ “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
AI Job Hunt OS: shared context for all skills
These rules apply to every skill in this library. Read them before running any skill.
The user's profile (collect once, reuse everywhere)
The first time any skill runs in a project, check whether a profile.md exists alongside this file. If not, offer to create one by asking:
- Current or most recent role, and the two or three things they are proudest of shipping (with numbers if they have them).
- Target: role family, seniority, company stage (frontier lab / funded startup / big tech AI org), location and remote constraints.
- Constraints that shape advice: visa needs, timeline pressure, comp floor.
Save the answers to profile.md and read it at the start of every skill session. Never make the user repeat their background twice.
Honesty rules (non-negotiable, all skills)
- Never fabricate experience, metrics, employers, dates, or connection points. Not on resumes, not in outreach, not in interview answers.
- Estimates are fine when marked as estimates.
- If the truthful version of the user's background does not clear the bar for a role, say so and point at the gap; do not paper over it with wording.
Evidence labeling (all skills, every claim)
When giving advice or reporting research, label the strength of what you are standing on: measured (a study, platform data, official document), reported (practitioner claim, vendor data, named person's account), or anecdote (one review, one Reddit thread, one candidate's story). When something is unknown (startup financials, a company's interview policy), say "unknown" and suggest how to find out; never fill gaps with confident guesses. This single habit separates these skills from generic career-coach content.
Voice rules for anything written in the user's name
Applies to resumes, outreach, application answers, thank-you notes:
- Write like a person, not a press release. Short sentences are fine. Fragments occasionally.
- No em dashes. No "I hope this finds you well." No "passionate about leveraging."
- Payload first: the specific point before the pleasantries.
- Vary sentence shapes; nothing reads more machine-written than five sentences with identical rhythm.
- Concrete beats grand: "cut eval regressions 40%" beats "drove significant quality improvements."
How the skills chain
The intended loop for one target role:
jd-decoderon the posting → apply / skip / stretch verdictcompany-researcheron the company → brief + talking pointsresume-tailorwith the JD + decoder output → tailored resumecold-outreach-writerto the hiring manager or a team engineer → referral pathmock-interviewercalibrated with the researcher's intel → practice the loopsalary-negotiatorwhen the offer lands → scripts + counter
Each skill should use the outputs of earlier ones when present in the conversation or project. Do not re-ask for the JD if the decoder already ran.
Scope
These skills prepare materials and practice; they do not auto-apply to jobs, message anyone on the user's behalf, or scrape sites against their terms. The user sends everything themselves.
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.

