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。

01 /

按职业浏览 Agent Skills

为不同岗位挑选的高频技能

全部职业
02 /

按使用场景浏览 Agent Skills

从你要完成的任务出发

全部场景
03 /

按创作者浏览 Agent Skills

按 GitHub Stars 排序的优秀创作者

全部创作者

全部 Skills 482556

muratovv
muratovv/ai-hats

article-outline

Outline before draft — the text's purpose, its Diátaxis type, audience, style forks and sections, agreed before a word of body text. Use when a text (article, doc page, guide, report) is being planned, or when a draft is about to start without an agreed type and reader.

内容创作者写作与文案
★ 3
muratovv
muratovv/ai-hats

api-evolution-checklist

Audit public-API/contract changes against Hyrum, Tesler, Leaky Abstractions, and Lehman laws. Use when changing a public API signature, response shape, or error contract, refactoring a shared interface used by two or more callers, drafting an interface ADR, or reviewing a PR touching api/, proto/, or OpenAPI specs.

开发工程师后端与接口
★ 3
muratovv
muratovv/ai-hats

ansible-ops

Ansible playbook management covering inventory, vault, syntax checks, and idempotency. Use when writing or running Ansible playbooks, configuring servers or deploying services, or managing secrets with Ansible Vault.

★ 3
muratovv
muratovv/ai-hats

agentic-topology-design

Choose a multi-agent topology and enforce structured contracts between sub-agents. Use when designing a flow with two or more sub-agents (code-review pipeline, deep research, incident triage, batch refactor), replacing a monolithic prompt agent showing attention dilution, or drafting an ADR for an agentic feature.

设计师视觉设计Skill与智能体开发
★ 3
muratovv
muratovv/ai-hats

agent-cli-guidelines

Design a CLI, gate or hook whose reader is another agent. Use when building or changing one, and when a tool of yours was ignored, misread, or worked around.

Skill与智能体开发
★ 3
muratovv
muratovv/ai-hats

adr-manager

Document architectural decisions using the Michael Nygard ADR format. Use when a significant architectural choice is made: picking a database, deciding on an API protocol, adopting a new pattern, or choosing between competing approaches.

后端与接口
★ 3
muratovv
muratovv/ai-hats

safety-guard

PreToolUse hooks that refuse destructive commands, ask before a shared-state write, block raw merges of task branches into a base branch outside the review flow, refuse a gate flag the agent writes for itself, keep the tracker writable only through `rack`, refuse the harness task list, and turn a consent-declared move into a question for the supervisor. Infrastructure — read it to diagnose a block, not to invoke it.

★ 3
muratovv
muratovv/ai-hats

worktree-isolation

Isolated development using git worktrees so the main branch stays clean. Use when starting any non-trivial task (execute state), doing parallel work on multiple tasks, making risky changes you might want to discard, or delegating to a sub-agent (automatic via ai-hats agent --isolation).

★ 3
muratovv
muratovv/ai-hats

tool-call-hygiene

Choose between Bash and dedicated Claude Code tools (Grep/Glob/Read/Edit) and edit files efficiently. Triggers on intent to grep/find/cat/sed/rg, multiple sequential reads, session-start context restore, codebase discovery, or multiple sequential Edits to one file.

★ 3
muratovv
muratovv/ai-hats

task-summary

Focused post-task summary covering architectural decisions, decision forks, pitfalls, and plan deviations. Use at `document → review`, before handing the card to the reviewer, on a task that failed, or when the supervisor asks what happened with a task.

★ 3
muratovv
muratovv/ai-hats

task-slicing

Wire a work breakdown into the tracker — depends_on edges, frontier, expand–contract for wide refactors. Use when filling plan.md's Steps section (routed from plan-gate), decomposing an epic into child tasks, or carving a mid-execute remainder into successor tasks.

律师与法务
★ 3
muratovv
muratovv/ai-hats

self-retrospective

Turn what went wrong into a systemic fix. Use when the supervisor asks for a retro, after a failed task (mandatory), after a task with backtracks, or when wrapping up a multi-task session. Not the automated audit that runs after a session closes.

★ 3
muratovv
muratovv/ai-hats

scope-guard

Enforce user-defined task boundaries to prevent scope creep and over-implementation. Use after receiving a task with explicit constraints ("only X", "don't do Y", "focus on Z"), before starting each new sub-step during execution, or when tempted to add helpful work not explicitly requested.

★ 3
muratovv
muratovv/ai-hats

safety-guard

PreToolUse hooks that refuse destructive commands, ask before a shared-state write, block raw merges of task branches into a base branch outside the review flow, refuse a gate flag the agent writes for itself, keep the tracker writable only through `rack`, refuse the harness task list, and turn a consent-declared move into a question for the supervisor. Infrastructure — read it to diagnose a block, not to invoke it.

★ 3
muratovv
muratovv/ai-hats

role-coherence-protocol

Audit a composed role against the user's project context for contradictions and interference, producing free-form findings with proposed fixes. Use when launched as a role-auditing role (role-auditor, role-judge, or sibling) whose first user message contains the target role's composed text plus the project context to audit against it.

★ 3
muratovv
muratovv/ai-hats

review-session

Single-pass review of one session run, producing a hats-reflect-session/v1 document (summary, observations, one hypothesis verdict per active HYP, proposal actions). Use when running as reflect-session or session-reviewer over a specific session id.

★ 3
muratovv
muratovv/ai-hats

review-proposal

Triage one improvement opportunity — vote on an existing PROP or create a novel one via `rack proposal`. Use when you spot an improvement (rule/skill/code/process/doc) during a session review, or are sweeping the open proposal inbox for triage.

★ 3
muratovv
muratovv/ai-hats

review-hypothesis

Vote on one active hypothesis (HYP-NNN) against session evidence — pick confirmed/refuted/inconclusive/n-a + a recommendation and persist via `rack hyp`. Use when sweeping active hypotheses during a session review (reflect-session, session-reviewer, or judge).

★ 3