craftkit
AGENTS.md0 starsChanged 46 days agoMITView on GitHub
CraftKit is a cross-agent toolkit for prompt and skill authoring. Skills live at skills/<name>/SKILL.md. Family section contracts: docs/skill-anatomy.md. This repo is type-1 — no spec/ here. improve(<skill>) · docs(<skill>) · refactor(<skill>) · chore:
In this repository5 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- craft-handoffSkill
- craft-promptSkill
- spec-charterSkill
- spec-grillSkill
The file
Files
AGENTS.md
CraftKit is a cross-agent toolkit for prompt and skill authoring. Skills live at skills/<name>/SKILL.md. Family section contracts: docs/skill-anatomy.md. This repo is type-1 — no spec/ here.
Invariants
- Markdown and plain text before code. No runtime, plugin system, or CLI.
- Tooling is Node (
"type": "module",node scripts/<name>.mjs, zero deps unless a dep costs less than maintaining the alternative). Python only for ML/eval pipelines that need it. - A skill must be usable from its
SKILL.mdalone. Spine text names the capability, not a provider's tool. Examples andguides/may name tools. - Explicit-only skills pair
disable-model-invocation: truewithagents/openai.yamlpolicy.allow_implicit_invocation: false. SKILL.mdhard ceiling is 500 lines (references/for the rest).npm run verifyfails above 220 lines or a 50-worddescription.- Multi-section artifacts use English XML tag names even when the body is not English.
- Paths in skill text are worktree-relative.
Commits
improve(<skill>) · docs(<skill>) · refactor(<skill>) · chore:
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.

