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
skill-creator
Guide for creating new skills in this system with proper frontmatter format. Use when adding new skills, updating skill templates, or reviewing skill structure. DO NOT USE FOR: general agent or instruction file customization (use agent-customization) or modifying existing skill logic/content.
session-startup
Automatic startup cleanup guard for new conversations. Use when deciding whether to run the session cleanup detector before the first reply, handling stale-state prompts, or preserving run-once startup semantics. DO NOT USE FOR: post-merge archival workflows (use post-pr-review) or general workflow troubleshooting outside startup detection (use process-troubleshooting).
session-memory-contract
Canonical operational contract for Agent Orchestra session-state survival and handoffs. Use when discovering where plan, design, review, startup, calibration, tracking, or plugin state belongs. DO NOT USE FOR: designing new storage mechanisms or writing rationale docs (use design-exploration or documentation-finalization).
safe-operations
Safe file-operation, issue-creation, and git/gh state-claim protocol for Agent Orchestra. Use when choosing workspace tools, avoiding unsafe file writes, creating GitHub issues under the workflow rules, reporting CI state, claiming a test failure is pre-existing, writing to a GitHub comment or PR body, acting on a cleanup detector's output, bumping the plugin version, concluding something is absent from a truncated search result, or re-checking cited claims after merging main into a branch. DO NOT USE FOR: application-level debugging or replacing agent judgment on whether work is in scope.
routing-tables
Concise reference for deterministic specialist dispatch, review mode routing, CE surface mapping, enum values, and gate criteria extracted from current orchestration agents. Use when an agent or script needs the canonical current routing data without re-deriving it from duplicated prompt prose. DO NOT USE FOR: taking orchestration decisions away from the calling agent, replacing judgment on uncovered cases, or storing mutable runtime state.
review-judgment
Reusable single-shot review judgment methodology for scoring prosecution and defense ledgers, verifying evidence, and emitting judge output, plus the parent-owned post-judge steps every review lane runs against that output — the close-out record's amendment rule (its single home on every lane) and the Post-Judge Disposition Gate. Use when ruling on review findings after prosecution and defense are available, and when a lane needs the close-out amendment trigger. DO NOT USE FOR: GitHub review intake routing, response-location policy for review responses, or fix execution ownership (keep those in Code-Review-Response.agent.md).
research-methodology
Evidence-driven research methodology for technical analysis and recommendation building. Use when gathering verified findings, cross-referencing internal and external sources, or converging multiple options into one recommended approach. DO NOT USE FOR: implementation work (use implementation-discipline) or debugging a live failure path (use systematic-debugging)
refactoring-methodology
Proactive refactoring workflow for touched files and nearby debt. Use when hunting extraction opportunities, evaluating duplication or oversized files, or improving structure without changing intended behavior. DO NOT USE FOR: adversarial defect review (use adversarial-review) or net-new feature implementation (use implementation-discipline)
property-based-testing
Incremental rollout policy for property-based testing that preserves readable example-based tests while adding invariant-focused randomized verification. Use when adding randomized testing, validating input ranges, or verifying mathematical properties. DO NOT USE FOR: example-based unit/integration tests or TDD workflow (use test-driven-development), React component tests (use ui-testing), or E2E browser tests (use webapp-testing).
project-references
Reference discoverability and loading methodology for Agent Orchestra. Use when project reference sidecars, index, or citation structure is relevant to the workflow or documentation. DO NOT USE FOR: general code navigation, symbol lookup, or when project reference structure is not relevant to the current task.
process-troubleshooting
Five-scenario workflow troubleshooting guide for diagnosing and fixing common orchestration failure patterns. Use when diagnosing premature implementation, agent confusion, validation gaps, documentation conflicts, or terminal stalls during a workflow session. DO NOT USE FOR: general code review (use code-review-intake), debugging application code (use systematic-debugging), or post-merge cleanup (use post-pr-review).
process-retrospective
Formalized skeleton for the process-retrospective frame port. Use when #348 ships the live producer and replaces this skeleton. DO NOT USE FOR: any current PR — applies-when is 'never' until #348 ships. See trigger-deferred-to and trigger-deferred-since below.
process-analysis
Reusable retrospective and process-analysis methodology for workflow reviews. Use when analyzing execution timelines, plan adherence, documentation conflicts, workflow efficiency, or skill-usage gaps after work completes. DO NOT USE FOR: CE Gate Track 2 defect triage format that is contract-owned by Process-Review.agent.md, or live implementation debugging before root-cause evidence exists (use systematic-debugging)
pre-commit-formatting
Backstop formatting workflow used during PR creation. Use when applying the final markdown and whitespace formatting gate before validation evidence capture. DO NOT USE FOR: per-step implementation formatting advice or PowerShell formatting rules outside this gate.
post-pr-review
Post-merge checklist for archiving, documentation, versioning, and release tagging. Use when completing post-merge cleanup, archiving tracking files, updating docs, or running the pre-merge strategic assessment (Step 6). DO NOT USE FOR: pre-PR readiness checks (use verification-before-completion) or processing GitHub review comments (use code-review-intake).
plugin-release-hygiene
Maintainer-side version-bump guardrail and Claude startup drift backstop guidance for plugin entry-point edits. Use when entry-point files change, when choosing patch/minor/major overrides, or when documenting/running the Claude plugin update surface. DO NOT USE FOR: CI release automation, registry publishing, or purely manual non-agent edit flows.
plan-authoring
Reusable implementation-plan authoring methodology. Use when running read-only discovery, drafting execution steps with CE Gate coverage, or preparing a plan for adversarial stress-testing and approval; also when writing acceptance criteria that name a consumer, a population, or a proof standard, when answering the vacuity question about a criteria set, when tagging a claim source-read or sample-inferred in an epistemic map, and when deciding where an instruction belongs so its executor actually reads it. DO NOT USE FOR: plan persistence, approval-policy enforcement, or direct implementation work (keep those in Issue-Planner.agent.md or use implementation-discipline)
persist-changes
Git-portable commit+push primitive for applied changes. Caller-parameterized; no Code-Conductor session flags. Use after a validated terminal step to commit staged fix files and push to the current branch's PR head remote. DO NOT USE FOR: new-PR creation (that is Code-Conductor Step 4 git push -u origin); force-push; or any scenario requiring git add -A.