release-guard

Core Rules (NEVER VIOLATE)

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "release-guard" with this command: npx skills add d-o-hub/rust-self-learning-memory/d-o-hub-rust-self-learning-memory-release-guard

Release Guard Skill

Core Rules (NEVER VIOLATE)

  • Releases ONLY from main after PR merge.

  • ALL CI jobs must COMPLETE + SUCCESS. NO queued jobs.

  • Verify branch protection requires status checks.

Activation Steps

  • PR Check: Ask for PR# if needed. Run gh pr view $PR_NUM --json state,baseRefName . Must: state=CLOSED, baseRefName=main.

  • Branch: git branch --show-current

  • gh repo view --json default_branch . Fail if not main.
  • CI: gh run list --branch main --limit 5 --json status,conclusion . ALL: completed + success. Log/screenshot.

  • Clean: git status (clean working dir).

Fail Response

🚫 BLOCKED: [Exact violation] Fix:

  • Merge PR to main
  • Wait CI: gh run list --branch main
  • git checkout main && git pull Re-run task.

Pass: Safe Release

  • Semver check: cargo semver-checks check-release --workspace (ADR-034)

  • Tag: vMAJOR.MINOR.PATCH (semantic).

  • Prefer: cargo release patch|minor|major (ADR-034)

  • Fallback: gh release create v1.2.3 --generate-notes --target main

  • Confirm before execute.

Examples

  • User: "Create release from develop PR #199" → BLOCK: Wrong branch.

  • User: "Tag v1.0.0 after PR #199" → Check PR/CI/branch → Proceed if pass.

Progressive disclosure: For CI details, see ci-reference.md if needed.

Best Practices Applied

  • Description: Keyword-rich for matching ("release", "tag").
  • allowed-tools: Read + Bash wildcards (gh/git CLI). Install gh.[2]
  • Structure: Essential in SKILL.md; link supports (add ci-reference.md for details).
  • Load/Test: Restart Claude Code. Ask "What Skills available?". Test: "Create release PR #199".

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

General

context-retrieval

No summary provided by upstream source.

Repository SourceNeeds Review
General

rust-async-testing

No summary provided by upstream source.

Repository SourceNeeds Review
General

build-rust

No summary provided by upstream source.

Repository SourceNeeds Review