claudius

bug-investigation

Use when investigating a reported bug, diagnosing a failure, or doing root-cause analysis — especially before concluding "not a bug", tracing a symptom, or explaining unexpected behavior. Preloaded on investigation agents.

Data & Documents Listed

check-pr-comments

Use to verify PR review comments are addressed in code. Optionally produces triage-compatible report.

DevOps & Infrastructure Listed

ci-dance

This skill should be used when the user says 'ci-dance', 'make the PR green', 'ship this and fix CI', 'push and handle reviews', or wants end-to-end PR pipeline automation.

coding-best-practices

Use when developing code. Universal rules for TDD, self-review, quality timing, review format, security. Preloaded on developers.

dependabot-merge

Bulk-process open dependabot PRs: audit each dependency, comment findings, merge if CI green, request rebase on conflicts or CI failures. Use to merge dependabot PRs, process dependency bumps, auto-merge bot PRs, or handle dependabot backlog.

Web & Frontend Listed

frontend-best-practices

Frontend best practices — TypeScript, React/Vue/Svelte, CSS, accessibility, testing. Use when writing, reviewing, or discussing frontend code.

git-and-github

Invoke for all git and gh commands, GitHub interactions. Solves git and gh access / permission denied issues.

Testing & QA Listed

go-best-practices

Go best practices — idioms, error handling, concurrency, testing patterns. Use when writing, reviewing, or discussing Go code.

grand-admiral

Multi-agent orchestration doctrine: spawning, worktree isolation, team coordination, scaling, recovery, programme management. Always loaded by coordinator agents that spawn, manage, and merge work from subagents.

grumpy-review

Parallel-agent code review for quality, security, dependencies, and docs. Use for reviews, audits, or quality assessments. Produces deduplicated severity-ranked report.

Data & Documents Listed

lessons-learned

Use when extracting learnings, saving lessons, or capturing reusable knowledge from the session. Also appropriate before presenting a plan, after notable events (bugs, wrong approaches corrected), and as final task when work is complete.

merge-base

Use for merging base into feature branch with conflict resolution.

Show all 27 bundled skills

push

Commit, push, and create PR. Auto-creates feature branch if on base. Use when user wants to commit and push, create a PR, ship work, send changes upstream, open a pull request, or publish a branch.

Testing & QA Listed

python-best-practices

Python best practices — PEP 8, type hints, testing, error handling, code quality tools. Use when writing, reviewing, or discussing Python code.

release

Bump version (SemVer 2.0), update changelog, commit, push, and create GitHub release. Auto-detects project tech stack (Rust, Python, JS/TS, Claude Code plugins, etc.). Args: major|minor|patch or auto-detect from commits. User-invocable only — agents must not invoke this skill autonomously.

report-format

Unified review report format for all finding-producing agents. Load when emitting or consuming review findings.

review-dependency

Use for security review of dependency updates — bumps, upgrades, or new dependencies.

review-loop

Use to autonomously handle peer review — request, wait, fix comments, push, repeat until resolved.

review-pr

Use to review a PR for code quality, security, and correctness.

rust-best-practices

Rust best practices checklists (Microsoft Pragmatic + Rust API Guidelines). Use when implementing, fixing, writing, reviewing, or discussing Rust code quality, API design, safety, or idioms.

API & Backend Listed

security-best-practices

OWASP-based secure programming practices. Use when writing or reviewing code handling auth, crypto, user input, secrets, or API endpoints. Consult proactively during reviews and planning.

severity

Use when rating findings in reviews, audits, and assessments. Preloaded on finding-producing agents.

triage-findings

Use for interactive browser-based triage of review findings. Only invoke when explicitly requested.

validate-findings

Coordinator-only LLM validation pass. Adds ai_assessment / ai_verdict / ai_verdict_confidence and, in the rare partial-producer case, re-estimates absent risk/impact/scope on a consolidated v3 report.

workflow-feature

Use for new projects, features, or major refactoring. Phases: Planning (Req→UX→Test Spec→Dev Plan) → Implementation → QA → Lessons Learned. Auto-retry on failure, unattended.

workflow-simplified

Use for bug fixes or small changes (≤200 lines). Same phase order as workflow-feature (Planning→Impl→QA→LL) with lighter ceremony. Auto-retry on failure, unattended.

workflow-trivial

Use for typos or single-line fixes (≤20 lines). Same mandatory phase order (Planning→Impl→QA→LL), minimal ceremony. Auto-retry on failure.