From review-pack
Run a structured code review of the current branch or a named PR using the review skill.
npx claudepluginhub s-hiraoku/claude-harnesses --plugin review-pack[PR number or 'branch']Run the `review` skill against $ARGUMENTS (or the current branch if empty). If the diff is large (>30 changed files), spawn multiple `code-reviewer` subagents in parallel — one per concern (correctness, security, performance, tests) — and combine their findings. Lead the response with the severity-ordered findings list, then a one-line verdict.
/reviewInvokes external AI CLIs (Gemini, Claude, Codex, OpenCode, Qwen, Cursor) to peer review specified phase plans, producing structured REVIEWS.md with feedback.
/reviewReviews staged changes or recent commits across five axes—correctness, readability, architecture, security, performance—producing categorized findings with file:line references and fixes.
/reviewRuns Codex code review on local git state (working tree or vs base branch). Supports --wait/--background, --base <ref>, --scope auto|working-tree|branch.
/reviewPerforms enhanced multi-LLM code review using providers like Codex, Gemini, Claude, and others, checks availability via bash, generates inline PR comments.
/reviewReviews HTML file for design anti-patterns, principles violations, and accessibility issues. Generates markdown report with status tables and recommendations.
/reviewReviews specified code scope via four specialists (quality, security, performance, architecture), producing summary, detailed findings, refactoring suggestions, prioritized action plan.
Share bugs, ideas, or general feedback.
Run the review skill against $ARGUMENTS (or the current branch if empty).
If the diff is large (>30 changed files), spawn multiple code-reviewer subagents in parallel — one per concern (correctness, security, performance, tests) — and combine their findings.
Lead the response with the severity-ordered findings list, then a one-line verdict.