api-test-pilot-skill
CLAUDE.md2 starsChanged 4 days agoMITView on GitHub
In this repositoryOne other file it gives its agents.
- api-testingSkill
The file
We split this CLAUDE.md into the rules it gives an agent
5 rules. About 0 are common, 3 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
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
- Very specific: line 9
Read PLAN.md first. Work phase by phase.
- Very specific: line 10
Test-type definitions live in TAXONOMY.md.
- Distinctive: line 17
Every generated test suite must be reproducible: print the exact command to rerun.
- Distinctive: line 21
Never print or log secrets.
- Distinctive: line 25
Keep files small and focused.
≈ “Common” is an estimate: at least 1 in 5 of the 27,641 CLAUDE.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
CLAUDE.md — api-test-pilot-skill
What this repo is
An Agent Skill (skills/api-testing/) that plans, generates and runs API tests
from an OpenAPI spec, Postman collection or curl examples. Goal: publish to
GitHub, then Claude and other AI skill marketplaces.
How to work here
- Read
PLAN.mdfirst. Work phase by phase; tick checkboxes as you finish. - Test-type definitions live in
TAXONOMY.md. Do not invent new types without updating it. - Record every scope/stack choice in
DECISIONS.md(date + reason). - Validate against the APIs in
eval-apis.mdbefore calling any phase done.
Skill design rules
SKILL.mdstays short (router only). Detail goes inreferences/*.md, loaded on demand.- Scripts must be dependency-light (stdlib where possible) and runnable offline.
- Every generated test suite must be reproducible: print the exact command to rerun.
- Safety gate (non-negotiable): load, stress, soak, spike, chaos and fuzz runs require the user to confirm the target is non-production and that they own or are authorized to test it. Never run these against an unconfirmed target.
- Never print or log secrets. Read credentials from env vars.
Conventions
- Markdown for docs, Python 3.11+ for scripts, k6 (JS) for load tests.
- Keep files small and focused; one test type per reference file.
- Conventional commits (
feat:,fix:,docs:).
Definition of done (per phase)
Checklist ticked in PLAN.md, eval run documented, DECISIONS.md updated, committed.
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.

