From 290e7c1f21083166cf3b19c4faa0996748851cc1 Mon Sep 17 00:00:00 2001 From: "github-actions[bot]" <41898282+github-actions[bot]@users.noreply.github.com> Date: Mon, 18 May 2026 11:51:39 +0000 Subject: [PATCH] docs: fix label name, testing guide, and workflow count in CONTRIBUTING Three documentation gaps surfaced by reviewing the agent-team dispatch input fix (d688265) against the existing docs: 1. README.md Uninstall: listed `state:in-progress` as the 7th label to delete, but that label doesn't exist. The real 7th label created by /install-agent-team is `agent-team:reviewed` (SKILL.md:80, reviewer-agent.md:64, catalog/agent-team/README.md:100). 2. CONTRIBUTING.md Testing: claimed "There is no automated test harness" but the repo has a full 3-tier suite under tests/ with CI integration (ci-tests.yml). Replaced with the actual tier-2/1/3 breakdown, run commands, and when to use each tier. 3. CONTRIBUTING.md Workflow files: said "seven dogfooded workflows" (the old count before b3f6341 trimmed the set). Now three gh-aw workflows are dogfooded. Also fixed a broken link to README.md#publishing (that section was removed in 83bdd43). Co-Authored-By: Claude Sonnet 4.6 --- CONTRIBUTING.md | 42 ++++++++++++++++++++++++++++++++++++++---- README.md | 2 +- 2 files changed, 39 insertions(+), 5 deletions(-) diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index 1c2721e..f666a81 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -74,18 +74,52 @@ Edit the relevant `SKILL.md` or data file. Test by running the skill locally wit ## Testing -There is no automated test harness for skills — they are instruction sets interpreted by Claude Code, not code with unit tests. The validation steps are: +The repo has a three-tier test suite under `tests/`. Run it with: + +```bash +bash tests/run-tests.sh +``` + +**Tier 2 — invariants** (fast, <1 s, always run): + +```bash +bash tests/test-invariants.sh +``` + +Filesystem and grep checks tied to past bugs. These run automatically in CI on every PR and push to `main`. Add a new invariant when you catch a regression in review — not preemptively. + +**Tier 1 — skill tests** (~4–5 min, run on changes to `skills/` or `catalog/`): + +```bash +bash tests/test-discover-workflows.sh +bash tests/test-install-workflow.sh +bash tests/test-install-agent-team.sh +``` + +Each test drives Claude Code with the plugin loaded and verifies the Skill tool was invoked and the response contains expected concepts. These also run in CI. + +**Tier 3 — E2E** (optional, 5–35 min, destructive against a real repo): + +```bash +bash tests/test-e2e-install-workflow.sh +bash tests/test-e2e-install-agent-team.sh +bash tests/test-e2e.sh # full spec→plan→impl→review pipeline +``` + +Run tier-3 manually against a scratch repo before opening a PR that changes agent-team workflow logic. See `tests/README.md` for details. + +**Manual validation steps** (always applicable): 1. **Load the plugin**: `claude --plugin-dir .` — confirm no startup errors. 2. **Run the skill manually**: invoke `/discover-workflows` or `/install-workflow` and walk through the flow. 3. **Validate lock files** (if you changed `.lock.yml` files): `gh aw validate` — safe, does not recompile. 4. **Check grep counts** (if you applied the OAuth tweak): see [skills/install-workflow/auth.md](skills/install-workflow/auth.md#step-4--verify-the-tweak-shape). -Never test by committing untested changes to `main`. The installed workflows run on push to `main`, so a broken install skill or a bad `.lock.yml` will trigger a live workflow run. +Never commit untested changes to `main`. The installed workflows run on push to `main`, so a broken install skill or a bad `.lock.yml` will trigger a live workflow run. ## Workflow files -The `.github/workflows/` directory contains seven dogfooded workflows. These are managed by `gh aw` — do not edit `.lock.yml` files by hand except to apply the OAuth tweak described in [skills/install-workflow/auth.md](skills/install-workflow/auth.md). +The `.github/workflows/` directory contains three dogfooded gh-aw workflows (`daily-repo-status`, `update-docs`, `weekly-research`) plus standard GitHub Actions workflows (`ci-tests.yml`, `agentics-maintenance.yml`, `copilot-setup-steps.yml`). The gh-aw workflows are managed by `gh aw` — do not edit `.lock.yml` files by hand except to apply the OAuth tweak described in [skills/install-workflow/auth.md](skills/install-workflow/auth.md). If a workflow `.md` source needs changing: @@ -112,4 +146,4 @@ Branch naming conventions: ## Publishing (maintainers only) -See the [Publishing section of the README](README.md#publishing) for the steps to submit the plugin to the Claude plugin registry. +See the [Releases](https://github.com/verkyyi/github-agent-runner/releases) page and the [claude-plugins.dev](https://claude-plugins.dev) listing for publishing context. The plugin is also listed on [ClaudePluginHub](https://claudepluginhub.com). diff --git a/README.md b/README.md index e18e51a..8f699ea 100644 --- a/README.md +++ b/README.md @@ -80,7 +80,7 @@ To remove workflows this plugin installed into your target repo: - `gh aw remove ` for each installed workflow (deletes both the `.md` source and the compiled `.lock.yml`), then commit the deletion. - `gh secret delete CLAUDE_CODE_OAUTH_TOKEN` — or `ANTHROPIC_API_KEY`, whichever path you used — to unset the auth secret. -- For `agent-team` specifically, also delete the seven labels: `gh label delete agent-team` plus `gh label delete state:` for each of `plan-needed`, `impl-needed`, `review-needed`, `done`, `blocked`, and `in-progress`. +- For `agent-team` specifically, also delete the seven labels: `gh label delete agent-team`, `gh label delete agent-team:reviewed`, and `gh label delete state:` for each of `plan-needed`, `impl-needed`, `review-needed`, `done`, and `blocked`. Nothing else is persisted — the plugin writes only to your target repo (under user approval) and holds no local state outside Claude Code's own plugin directory.