skills
AGENTS.md1 starsChanged 3 months agoMITView on GitHub
Categories are engineering, open-source, review, productivity. A skill lives in exactly one category; if it plausibly belongs in two, the category is wrong. Frontmatter carries name (must equal the directory name) and description. The description is what an agent reads when deciding whether to load the skill, so it states what the skill does and when to use it — not what topic it is about. Keep it under 1024 characters. The body is instructions addressed to whoever is doing…
In this repository45 other files it gives its agents.
- claude-handoffSkill
- codebase-designSkill
- code-reviewSkill
- diagnosing-bugsSkill
- domain-modelingSkill
- git-guardrails-claude-codeSkill
- grillingSkill
- grill-meSkill
- grill-with-docsSkill
- handoffSkill
- implementSkill
- improve-codebase-architectureSkill
- loop-meSkill
- migrate-to-shoehornSkill
- prototypeSkill
- researchSkill
- resolving-merge-conflictsSkill
- scaffold-exercisesSkill
- setup-pre-commitSkill
- setup-skillsSkill
- setup-ts-deep-modulesSkill
- tddSkill
- teachSkill
- to-questionnaireSkill
- to-specSkill
- to-ticketsSkill
- triageSkill
- wait-whatSkill
- wayfinderSkill
- wizardSkill
- writing-beatsSkill
- writing-for-agentsSkill
- writing-fragmentsSkill
- writing-shapeSkill
- eslint-rule-authorSkill
- minimal-reproSkill
- rn-list-performanceSkill
- first-contributionSkill
- pr-that-mergesSkill
- survive-reviewSkill
- triage-issueSkill
- contribution-logSkill
- release-checklistSkill
- reproducibility-checkSkill
- review-research-softwareSkill
The file
Files
Working in this repository
Layout
skills/<category>/<skill-name>/SKILL.mdCategories are engineering, open-source, review, productivity. A skill
lives in exactly one category; if it plausibly belongs in two, the category is
wrong.
Writing a skill
Frontmatter carries name (must equal the directory name) and description.
The description is what an agent reads when deciding whether to load the skill,
so it states what the skill does and when to use it — not what topic it is
about. Keep it under 1024 characters.
The body is instructions addressed to whoever is doing the work. Prefer:
- Ordered procedures over prose essays.
- Named failure modes over general advice.
- The reason a step exists, where it is not obvious. A step without a reason is the first one skipped.
Cross-reference other skills with [[skill-name]].
Before committing
node scripts/validate.mjsDoes 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.