agentleFS
Sign inStart free
quantumquirkxyzquantumquirkxyz/skills-quirk/AGENTS.md

skills-quirk

AGENTS.md0 starsChanged 43 days agoMITView on GitHub

Instructions for coding agents working in this repository. This file complements README.md (for humans) with the operational rules an agent needs.

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

4 rules. About 0 are common, 2 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    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
  • could be automatic: a setting could enforce it
  1. Distinctive:

    Keep changes inside the declared scope of the task, skill, or ticket you are executing.

    line 22
  2. Very specific:

    Never merge with unresolved review findings.

    line 24
  3. Very specific:

    Keep generated artifacts in sync with .agents/skills/.

    line 36
  4. Distinctive:

    Do not commit unless explicitly asked.

    line 37

≈ “Common” is an estimate: at least 1 in 5 of the 36,368 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

AGENTS.md

Instructions for coding agents working in this repository. This file complements README.md (for humans) with the operational rules an agent needs.

Repository shape

  • Canonical skills live in .agents/skills/ as <domain>/<skill>/SKILL.md directories. They are the only source of truth for skill content.
  • .claude/skills/ is a generated compatibility view of relative symlinks. It is committed intentionally so Claude-style consumers can discover the same canonical skills, but it is regenerated by bundle tooling; never create, edit, or delete entries there by hand.
  • skills-lock.json records per-skill SHA-256 hashes. Editing a SKILL.md requires re-locking the corresponding hash, or the semantic audit fails.
  • Project vocabulary is authoritative in CONTEXT.md: quirk Skills, quirk Method, canonical skill, compatibility view, provenance. Use these terms; do not import vocabulary from unrelated repositories.
  • .agents/AGENTS.md covers the bundle directory itself; this file covers the repository.

Work-item governance

  • Work-item governance lives in docs/reference/agents/. Read docs/reference/agents/index.md first before routing specs, tickets, project boards, or publication flows; work-item-router enforces the same preflight.
  • Preserve the canonical work-item metadata shape from docs/reference/agents/work-item-format.md whenever specs, tickets, or PR bodies are produced.
  • Triage labels and tracker configuration follow docs/reference/agents/triage-labels.md and docs/reference/agents/issue-tracker.md.

Method

  • The bundle's own quirk Method applies to work done in this repo: artifact-first design, state-machine workflows, and shift-left quality gates.
  • Keep changes inside the declared scope of the task, skill, or ticket you are executing. Surface blockers instead of broadening scope.
  • Reference issues and tickets by name, not by bare ids.
  • Never merge with unresolved review findings. Findings from review-pr separate into Standards and Spec axes; repair them with an explicit plan via plan-review-fixes and implement-review-fixes, then re-review.

Commands

  • Full local gate (run before claiming any work done): node .agents/skills/platform/check-all.mjs
  • Tests: node --test tests/*.test.mjs
  • Regenerate projections (CATALOG.md, llms.txt, .claude-plugin/marketplace.json, and committed generated compatibility views): node scripts/sync-registry.mjs --write
  • Version alignment is enforced by tests/version-consistency.test.mjs (VERSION, package.json, registry.yaml, README badge).

Style

  • Canonical skills, prompts, and this file are written in English.
  • Keep generated artifacts in sync with .agents/skills/; never hand-edit them.
  • Do not commit unless explicitly asked.

Download now (ZIP)

  • 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 28,693 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 repository2 other files it gives its agents.

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

Reviews can't be read right now.

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.

Check your setup, free