agentleFS
radra23radra23/otel-as-code-plugin/.claude/skills/validate-plugin/SKILL.md

validate-plugin

Skill1 starsChanged 47 days agoApache-2.0View on GitHub

Run the local validation suite for the otel-as-code plugin before committing or opening a PR — hook tests, fixture integrity, plugin.json manifest checks, shell syntax, and terraform-validate on every committed snapshot. Covers the GitHub Actions CI checks plus extra local-only sanity.

5 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 1 is distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 4 edited versions we read keep these.

  • 1

    Distinctive

    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
  1. Distinctive:

    Run the project's CI checks locally (plus a couple of local-only sanity checks) and report a concise PASS/FAIL per group plus an overall …

    line 9
  2. Very specific:

    If the claude CLI isn't available: confirm both .claude-plugin/plugin.json and .claude-plugin/marketplace.json are valid JSON, plugin.json …

    line 24
  3. Very specific:

    Run the snapshot loop under bash -eo pipefail to match CI semantics (the local shell is zsh, which behaves differently on glob no-match).

    line 33

“Common” is measured: kept by at least 6 in 10 of the 4 edited versions of this file we could read. “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
SKILL.md

Validate the otel-as-code plugin

Run the project's CI checks locally (plus a couple of local-only sanity checks) and report a concise PASS/FAIL per group plus an overall verdict. On failure, show the failing command's output.

1. Hook tests (CI: lint-hooks)

Iterate every suite under tests/hooks/ — do NOT hardcode the list, so new tests run automatically:

for t in tests/hooks/*.test.sh; do echo "== $t =="; bash "$t" || exit 1; done

2. Fixture integrity (CI: validate-fixtures)

  • nodejs-greenfield: package.json, index.js, Dockerfile exist AND package.json has NO @opentelemetry dependency.
  • python-greenfield: pyproject.toml, app.py, Dockerfile exist AND pyproject.toml has NO opentelemetry dependency.
  • nodejs-brownfield: tracing.js exists AND still contains the three seeded violations — service.name as a span attribute, http.method, and un-namespaced orderId — which the lint tests assert against.

3. Plugin manifest & structure (CI: validate-plugin-manifest)

  • claude plugin validate . --strict — must pass. Validates plugin.json (metadata-only), marketplace.json, and the auto-discovered components (commands/*.md, agents/*.md, skills/<name>/SKILL.md, hooks/hooks.json).
  • If the claude CLI isn't available: confirm both .claude-plugin/plugin.json and .claude-plugin/marketplace.json are valid JSON, plugin.json declares NO inline component arrays, and the convention dirs/files exist.

4. Shell sanity (local-only — no CI equivalent)

  • bash -n on every hooks/*.sh and tests/**/*.sh.

5. Terraform snapshots (CI: validate-terraform)

For each tests/snapshots/<vendor>/main.tf.snap:

  • Copy it to a temp dir as main.tf, then terraform init -backend=false -input=false + terraform validate.
  • terraform is NOT installed locally — download a pinned binary (e.g. 1.7.x from releases.hashicorp.com) to a temp path and use that. (CI installs it via hashicorp/setup-terraform.)
  • Run the snapshot loop under bash -eo pipefail to match CI semantics (the local shell is zsh, which behaves differently on glob no-match).

Report

One line PASS/FAIL per group (hooks / fixtures / manifest / shell / terraform) and an overall verdict. This is the gate to clear before committing.

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.