yuga-hashimoto/open-test-pilot/publish-tests
Claude Code and agent skill by yuga-hashimoto/open-test-pilot
Trust tier 3
- community
Source check pending
Checking that the SKILL.md file is still in the repository.
Install
npx vskill@latest install yuga-hashimoto/open-test-pilot/publish-testsSKILL.md
Publish tests
Turn a reviewed test (see review-tests) into a proposal a human can approve — never a direct change to a protected branch, and never "done" without an explicit human decision recorded.
Always show the diff
Before proposing anything, present both diffs so the reviewer can see cause and effect together:
- Manifest diff — the YAML change (
testpilot manifest diff <before> <after>locally, or a plain git diff of the Manifest file). This is the actual source-of-truth change. - Generated-code diff — the resulting Playwright/Appium TypeScript diff. Since generated code is deterministic, this diff should be fully explained by the Manifest diff; if it isn't (e.g. unrelated formatting churn), investigate before proposing.
Local / repository mode
- Create a dedicated branch for the change — never commit directly to
mainor another protected branch. - Open a Pull Request (Draft, if the GitHub integration is configured) containing both diffs and a summary of what was analyzed, designed, generated, run, and (if applicable) repaired, with links to run evidence.
- If the change addresses a suspected product defect rather than a test fix, say so explicitly in the PR description and do not present it as a routine test update.
Team mode
- Use the GitHub integration to open a dedicated branch and Draft PR (never push straight to a protected branch); register the result with the
pull_request_register { organizationId, url }MCP tool so the platform tracks it against the run/test. - Always also call
change_request_create { organizationId, title, description }— even when a PR exists — so the proposal appears in the dashboard's AI view. The PR is the code artifact; the change request is the human approval gate. Both should reference the same run evidence. - Treat the proposal as pending until a human moves the change request to
approved(seehandle-change-request) or the PR is merged by a human reviewer.rejectedmeans stop — do not resubmit the same change without new information.
Hard rules
- Never push to a protected branch (
main, release branches) directly, in either mode. - Never merge your own proposal or bypass required review.
- Never omit the generated-code diff — a reviewer approving a Manifest change without seeing the resulting code is approving blind.
- If secrets, credentials, or tenant data would appear in the diff, stop and redact/reference instead of publishing (see
generate-manifest's secrets rules).
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingpackages/claude-code-plugin/skills/publish-tests/SKILL.md
Popularity
- Repository stars
- 1
- 7-day trend score
- 0
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/yuga-hashimoto/open-test-pilot/publish-tests)