agentleFS
leosiminileosimini/keelokit/AGENTS.md

keelokit

AGENTS.md2 starsChanged 10 days agoMITView on GitHub

This repo is Keelokit itself: a Claude Code plugin and the Copier template it generates projects from. The projects' own agreement is template/AGENTS.md.jinja; don't mix the two. This repo runs on its own harness (.keelokit/, adopted 2026-09-28): docs/context/, backlog/, the doctor in CI, the guard in your Claude sessions. Once per clone, git config core.hooksPath .githooks turns on the git hooks (the guard before a commit, scripts/verify.sh before a push). /keelokit says where things stand; stories are built with /keelokit:build-story,…

The file

We split this AGENTS.md into the rules it gives an agent

2 rules. About 0 are common, 0 are distinctive, and 2 are 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.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read docs/releasing.md before anything that touches versions, tags or the release branch.

    line 36
  2. Very specific:

    Never change version in .claude-plugin/plugin.json, push to release or create tags by hand.

    line 40

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
AGENTS.md

Keelokit — working agreement for this repo

This repo is Keelokit itself: a Claude Code plugin and the Copier template it generates projects from. The projects' own agreement is template/AGENTS.md.jinja; don't mix the two.

Layout

.claude-plugin/   plugin.json (name, version, description), marketplace.json, icon.svg
skills/ agents/ workflows/ hooks/ references/   the plugin
template/ copier.yml                 the project template (rendered from `template/`)
tests/            unit tests for the guard, doctor, dashboard, workflows and verify.sh's scope
scripts/          test-template.sh (generates projects and runs their checks), release.sh
docs/             design.md (how the pieces fit), releasing.md (how a version ships)

Commands

WhatCommand
Everything CI checks, locally (the pre-push hook runs it)scripts/verify.sh
Guard, doctor, dashboard, workflow and verify testspython3 -m unittest discover -s tests (test_verify.py needs pnpm, test_first_commit.py uv, the dashboard layout test playwright's Chromium)
Plugin manifestsclaude plugin validate . --strict
A generated project, end to endscripts/test-template.sh '["api","web"]' (see its header)

Keelokit on Keelokit

This repo runs on its own harness (.keelokit/, adopted 2026-09-28): docs/context/, backlog/, the doctor in CI, the guard in your Claude sessions. Once per clone, git config core.hooksPath .githooks turns on the git hooks (the guard before a commit, scripts/verify.sh before a push). /keelokit says where things stand; stories are built with /keelokit:build-story, one at a time. None of this ships in the plugin: scripts/package.sh leaves it out.

Releases

Read docs/releasing.md before anything that touches versions, tags or the release branch. In short:

  • Work lands on main. User-facing changes get a line under ## Unreleased in CHANGELOG.md.
  • Never change version in .claude-plugin/plugin.json, push to release or create tags by hand. A release is scripts/release.sh X.Y.Z, and only when the user asks for one.
  • The Claude plugin directory tracks release and scans the whole repo. A new scan finding gets a real fix or a line in the submission notes, never a rewording just to get past the scanner.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.