kickoff-ai
AGENTS.md1 starsChanged 3 months agoMITView on GitHub
Tool-neutral operating contract for any coding agent. The binding foundation is dev/base/CONSTITUTION.md; this file agrees with CLAUDE.md and never contradicts either. Setting up, checking, repairing or migrating a Mac with this repository → you are the OPERATOR. Read docs/12-OPERATOR-PLAYBOOK.md and follow it exactly. It defines what you may run unasked, what needs a confirmation first, and what you never run. Nothing else in this file applies to you. Changing this repository's own code or documentation → you are a CONTRIBUTOR.…
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.
In this repository5 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- contextSkill
- project-stateSkill
- session-startSkill
- session-stopSkill
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 19
Read dev/base/CONSTITUTION.md, then this repo's BIBLE.md, then HANDOFF.md (newest entry).
≈ “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
AGENTS.md — kickoff-ai
Tool-neutral operating contract for any coding agent. The binding foundation is
dev/base/CONSTITUTION.md; this file agrees with CLAUDE.md and never contradicts either.
Which role are you in? Decide this first.
Setting up, checking, repairing or migrating a Mac with this repository → you are the
OPERATOR. Read docs/12-OPERATOR-PLAYBOOK.md and follow it
exactly. It defines what you may run unasked, what needs a confirmation first, and what you never
run. Nothing else in this file applies to you.
Changing this repository's own code or documentation → you are a CONTRIBUTOR. Continue below.
Boot sequence (do this before editing anything)
- Read
dev/base/CONSTITUTION.md, then this repo'sBIBLE.md, thenHANDOFF.md(newest entry). - Inspect the current repo structure.
- Detect, don't assume: package manager · framework · routing model · styling system · lint/test/build commands.
- Produce a concise implementation plan before editing files.
- Preserve existing architecture unless there is a clear, stated reason to change it.
Gates — don't rely on memory, run them
- Python repo:
./scripts/gate.shmust printGATE: PASS. - Web repo:
npm run verify:ci(typecheck · lint · budget ratchets · tests · build). - Budget lints (
lint:tokens,lint:loc) are ratchets with committed baselines — fix violations and update the baseline in their own commit; never raise a budget to silence a regression.
If docs and code disagree
Trust the code, update the docs, and note the contradiction in your final summary.
Evidence protocol (any substantive finding)
Claim → Evidence (file, exact line, short verbatim quote, link) → Interpretation → Counter-check → scoped action item. Missing facts are marked as an open decision, never improvised.
Definition of done
A change is complete only when: it aligns with CONSTITUTION.md + BIBLE.md; lint/build/type
checks are run where available and green; the gate passes; changed files are summarized; and
remaining risks or assumptions are stated.
Commits
Conventional Commits, one concern per commit, documenting what and why. Breaking changes go in
CHANGELOG.md.
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.

