GITHUB AGENT SKILLS / 技能收录
给你的 AI 助手
装上专业技能
我们从 GitHub 持续收录、审核并分类优质的 Agent Skills,按职业与使用场景整理,复制一行命令即可安装到 Claude Code、Cursor 等智能体。
按职业浏览 Agent Skills
为不同岗位挑选的高频技能
按使用场景浏览 Agent Skills
从你要完成的任务出发
按创作者浏览 Agent Skills
按 GitHub Stars 排序的优秀创作者
Microsoft
@microsoftopenclaw
@openclawAnthropic
@anthropicsOpenAI
@openaiThe Apache Software Foundation
@apacheGitHub
@githubAutoGPT
@Significant-GravitasGarry Tan
@garrytan全部 Skills 482556
bmad-editorial-review-prose
Clinical copy-editor that reviews text for communication issues. Use when user says review for prose or improve the prose
gem-hunt
Use when the user wants to DISCOVER high-value improvements to their OWN agentic workspace/toolkit - new hooks, behaviors, gems, skills, MCPs, or frontier-harness patterns - via a wide multi-lens agent fan-out, then a consolidated adopt/prune decision. The named "agentic-perfection sweep" / "extensive gem search". Triggers: "find more gems", "what are we missing vs other tools".
bmad-shard-doc
Splits large markdown documents into smaller, organized files based on level 2 (default) sections. Use if the user says perform shard document
bmad-review-edge-case-hunter
Walk every branching path and boundary condition in content, report only unhandled edge cases. Orthogonal to adversarial review - method-driven not attitude-driven. Use when you need exhaustive edge-case analysis of code, specs, or diffs.
writing-skills
Use when creating new skills, editing existing skills, or verifying skills work before deployment
bmad-review-adversarial-general
Perform a Cynical Review and produce a findings report. Use when the user requests a critical review of something
skill-tree-reconcile
Use when asked to sync skills, reconcile skill trees, deploy skills to live, choose which tree wins, compare live vs vault vs repo, run sync-from-live, or repair drift - classifies differences and applies a human-reviewed direction.
bmad-party-mode
Orchestrates group discussions between all installed BMAD agents, enabling natural multi-agent conversations. Use when user requests party mode.
skill-stocktake
Use when auditing Claude skills and commands for quality. Supports Quick Scan (changed skills only) and Full Stocktake modes with sequential subagent batch evaluation.
bmad-init
Initialize BMad project configuration and load config variables. Use when any skill needs module-specific configuration values, or when setting up a new BMad project.
skill-scout
Search local, marketplace, GitHub and web sources before creating a new skill. Use when the user wants to find, fork or build one. NOT for authoring it (use writing-skills).
bmad-index-docs
Generates or updates an index.md to reference all docs in the folder. Use if user requests to create or update an index of all files in a specific folder
packaged-artifact-smoke
Use when you need to prove a BUILT deliverable actually works end-to-end, not just that its source tree is CI-green. Runs 3-5 golden journeys against the PACKAGED artifact (the built Electron binary via `_electron.launch`, or `pip install dist/*.whl` in a fresh venv then invoke the installed CLI), each ending in a DETERMINISTIC on-disk postcondition probe (ffprobe, or file existence + size).
bmad-help
Analyzes current state and user query to answer BMad questions or recommend the next skill(s) to use. Use when user asks for help, bmad help, what to do next, or what to start with in BMad.
mcp-fleet-curation
Use when plugins, skills, or MCP servers keep REAPPEARING after you disable them, when de-bloating the Claude Code catalogue, or when the per-turn tool/token count creeps back up. The permanent cure - disable at USER scope (claude plugin disable name@mkt --scope user), NOT the UI toggle (which writes project scope and returns when the cwd changes) - plus periodic mcp_doctor audit.
bmad-editorial-review-structure
Structural editor that proposes cuts, reorganization, and simplification while preserving comprehension. Use when user requests structural review or editorial review of structure
bmad-editorial-review-prose
Clinical copy-editor that reviews text for communication issues. Use when user says review for prose or improve the prose
workflow-authoring
Use when writing or debugging a saved Claude Code Workflow (a .js/.mjs orchestration script under ~/.claude/workflows that uses the injected agent/parallel/phase globals to fan out subagents). Covers the exact runtime API, the args-as-string gotcha, the banned Date.now() non-determinism, worktree isolation for file-mutating agents, one-level nesting, a schema on every fan-in, null-tolerant barriers via filter(Boolean), per-work-unit pipelines so one slow agent cannot deadlock the cohort, cache-cliff-aware fan-out, the typed Codex output-schema landmine when MCP tools are active, and the SDK context-editing / task-budgets / canUseTool notes. Also the map to the saved schema-typed gate Workflows plan-gate and design-gate and the mechanical plan-lint stage-0. Triggers include author a workflow, write a saved workflow, fan out subagents, my parallel workflow deadlocks, workflow returns empty, schema-typed reviewers, or codify a gate as a workflow. NOT for single ad-hoc Task subagents (spawn a Task directly) or for the metaswarm phase pipeline itself.