jev-job-search
AGENTS.md368 starsChanged 5 days agoMITView on GitHub
This file is read by Codex, Cursor, Gemini CLI and other agents that look for AGENTS.md. Claude Code reads CLAUDE.md. The rules are the same for every agent and they live in one place: Read CLAUDE.md now, all of it, before you do anything in this repository. Where it says "Claude" or "you", it means you. The section "What never changes" is not advice: it is what keeps a person's applications truthful and their passwords out of your sight. The…
- Reads credentials
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 33,315 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
The file
We split this AGENTS.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 43
If data/profile.json does not exist, the person has not set the tool up.
≈ “Common” is an estimate: at least 1 in 5 of the 42,283 AGENTS.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
jev-job-search: for Codex and any other coding agent
This file is read by Codex, Cursor, Gemini CLI and other agents that look for
AGENTS.md. Claude Code reads CLAUDE.md. The rules are the same for every
agent and they live in one place:
Read CLAUDE.md now, all of it, before you do anything in this repository.
Where it says "Claude" or "you", it means you. The section "What never
changes" is not advice: it is what keeps a person's applications truthful and
their passwords out of your sight.
Running the tool for a person
The person does not type npx. They say what they want, and you run the CLI
from the repository root with npx jev <command>. Each workflow is written
down as a skill:
| They say | Read and follow |
| set this up, get me started | .claude/skills/setup/SKILL.md |
| find jobs, refresh the queue | .claude/skills/discover/SKILL.md |
| apply, start applying | .claude/skills/apply/SKILL.md |
| I typed the code, I signed in, continue | .claude/skills/resume/SKILL.md |
| where am I, what is waiting | .claude/skills/status/SKILL.md |
| open the dashboard | .claude/skills/report/SKILL.md |
| change my profile, an answer was wrong | .claude/skills/profile/SKILL.md |
| tailor my resume, write a cover letter | .claude/skills/tailor/SKILL.md |
| Workday, accounts, sign in | .claude/skills/accounts/SKILL.md |
| run it every day | .claude/skills/daily/SKILL.md |
| did anyone reply | .claude/skills/inbox/SKILL.md |
| add my GitHub projects | .claude/skills/expand/SKILL.md |
| add this job board | .claude/skills/add-source/SKILL.md |
Codex finds the same skills on its own under .agents/skills, which is a link
to .claude/skills: type $apply, or /skills to list them. On a checkout
without symbolic links (Windows without developer mode), open the files in
the table directly.
Where a skill says "Claude Code", read your own name. Where it says to show a
command in a bash block for the person to run, do exactly that: those are
the commands that sign in at an employer, and they are never yours to start.
If data/profile.json does not exist, the person has not set the tool up.
Offer the setup skill before anything else. npx jev doctor says what is in
place and what to do next.
Writing with your own model
By default the tool's writer is Claude. A person who works with Codex or
another model sets WRITER_BACKEND in .env (see .env.example and
docs/SETUP.md). Do not set it for them without asking: it decides which
company reads their profile.
Before pushing
npx tsc --noEmit && npx vitest run && npm audit --omit=devCommits follow Conventional Commits and carry no AI attribution of any kind.
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.

