agentleFS
AlperenEvciAlperenEvci/claude-code-development-harness/.claude/skills/harness-fixture-smoke-test/SKILL.md

harness-fixture-smoke-test

Skill1 starsChanged 2 months agoMITView on GitHub

Exercise the setup pipeline end to end against a disposable fixture repository outside this project, to prove that inspection, rendering, validation, and dry-run installation still work together. Invoke explicitly after changing templates, the renderer, the validator, or the installer.

The file

Files
SKILL.md

harness-fixture-smoke-test

  • The unit suite covers the scripts in isolation. This procedure proves the pipeline composes. Never run it against a real project.
    1. Create a scratch fixture directory outside this repository. Never use the project root as a fixture target.
    1. For an existing fixture, add a small but real shape: a manifest, a source directory, and a README. For a new fixture, leave it empty or add only a planning README.
    1. Run the inspector against the fixture and read project_state from the JSON. Confirm it classifies the fixture the way you intended.
    1. Start from a documented profile under examples/ rather than writing one from scratch, then adjust only the fields under test.
    1. Render into a staging directory with render_harness.py, then run validate_harness.py against the output. Both must pass before you go further.
    1. Run the generated install-harness.sh --target <fixture> --dry-run first. Read the NEW / IDENTICAL / CONFLICT / BLOCKED classification and confirm it matches what you expect.
    1. Run --apply-new-only, then run check_installed.py against the fixture.
    1. To exercise the conflict path, pre-create a file that the payload also provides, re-run the dry run, and confirm it is reported as CONFLICT and is not overwritten.
    1. Delete the fixture and the staging output when finished.
  • Report what you actually observed at each step, including any step you could not complete on this platform.

User-supplied context:

$ARGUMENTS

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.