Browse skills

Evaluate, fix, and reply to GitHub pull request review comments. Use when the user asks to "resolve PR comments", "fix review comments", "address PR feedback", "handle review comments", "address review feedback", "respond to PR comments", or "address code review".

Check API, library, and framework usage in code against official documentation and installed skill knowledge. Flags deprecated APIs, incorrect method signatures, wrong parameter types, version-incompatible patterns, and best-practice violations. Use when the user asks to "review API usage", "check API usage", "verify against docs", "check library usage", "validate API calls", "check against documentation", or "check for deprecated APIs".

Full code review: launches `/review-test-coverage`, `/review-correctness`, `/review-security`, `/review-quality`, `/review-api-usage`, and `/peer-review-code` in parallel and returns combined findings. Use when the user asks to "review my code", "full code review", "review my changes", or wants a comprehensive code review.

Analyze code for bugs and correctness problems using strict determination criteria. Returns structured findings without applying fixes. Use when the user asks to "review correctness", "scan my code for bugs", "find bugs in my changes", "look for bugs", "check for bugs", or "run a correctness review".

Detect package managers and discover outdated or vulnerable dependencies. Returns structured findings without upgrading. Use when the user asks to "review dependencies", "check for outdated packages", "check dependencies", "scan dependencies", or "dependency review".

Review an implementation plan: launches an internal plan review and `/peer-review-plan` in parallel and returns combined findings. Use when the user asks to "review my plan", "review this plan", "check my plan", "critique my plan", or wants plan feedback before implementation.

Review a pull request by fetching PR comments and running a comprehensive code review with evaluation. Use when the user asks to "review PR", "review pull request", "review this PR", "check PR before merging", or "full PR review".

Tones down visually aggressive or overstimulating designs, reducing intensity while preserving quality. Use when the user mentions too bold, too loud, overwhelming, aggressive, garish, or wants a calmer, more refined aesthetic.

Review a prompt plan against its source spec: launches an internal prompt plan review and `/peer-review-prompt-plan` in parallel and returns combined findings. Use when the user asks to "review my prompt plan", "review my prompts", "check my prompt plan", "critique my prompts", or wants prompt plan feedback before implementation.

Analyze code for cross-file quality issues: duplicated logic, architectural inconsistencies, and abstraction opportunities. Returns structured findings without applying fixes. Use when the user asks to "review quality", "check for duplication across files", "find architectural inconsistencies", "cross-file quality review", or "review code quality".

Analyze code for security vulnerabilities, insecure patterns, and security regressions using strict determination criteria. References the project threat model when available. Returns structured findings without applying fixes. Use when the user asks to "review security", "check for security issues", "find vulnerabilities in my changes", "run a security review", "analyze my code for security", "security audit", or "security scan".

Review a specification document: launches an internal spec review and `/peer-review-spec` in parallel and returns combined findings. Use when the user asks to "review my spec", "review this spec", "check my spec", "critique my spec", or wants spec feedback before planning.

Detect what dev tooling infrastructure a project has and flag gaps across linters, formatters, pre-commit hooks, test runners, and CI/CD pipelines. Returns structured findings without applying changes. Use when the user asks to "review tooling", "check project tooling", "what tooling is missing", "review dev infrastructure", or "tooling audit".

Extract lessons from the current session and route them to the appropriate knowledge layer (project AGENTS.md, auto memory, existing skills, or new skills). Use when the user asks to "self-improve", "distill this session", "save learnings", "update CLAUDE.md with what we learned", "capture session insights", "remember this for next time", "extract lessons", "update skills from session", or "what did we learn".

Launch the app and hands-on verify that it works by interacting with it. Use when the user asks to "smoke test", "test it manually", "verify it works", "try it out", "run a smoke test", "check it in the browser", or "does it actually work". Not for unit/integration tests.

Stage files, create a commit, and push in one step. Use when the user asks to "stage commit and push", "add commit and push", "commit and push", or "commit and push my changes".

Update the Unreleased section of CHANGELOG.md based on current changes. No-op if CHANGELOG.md does not exist. Use when the user asks to "update changelog", "add to changelog", "update the changelog", "changelog entry", "add changelog entry", or "log this change".

Create and manage datasets on Hugging Face Hub. Supports initializing repos, defining configs/system prompts, streaming row updates, and SQL-based dataset querying/transformation. Designed to work alongside HF MCP server for comprehensive dataset workflows.

Build AI chat interfaces using ai-elements components — conversations, messages, tool displays, prompt inputs, and more. Use when the user wants to build a chatbot, AI assistant UI, or any AI-powered chat interface.

强大的信息获取与浏览器自动化工具。通过浏览器 + 用户登录态，获取公域和私域信息。可访问任意网页、内部系统、登录后页面，执行表单填写、信息提取、页面操作。支持 site 系统（36 平台 103 命令一键调用）、带登录态的 fetch、网络请求拦截与 mock、操作录制等高级功能。