215 lines
5.6 KiB
Markdown
215 lines
5.6 KiB
Markdown
# AI技能库目录
|
||
|
||
`AI技能` `智能体` `可复用组件` `技能管理`
|
||
|
||
# 技能库
|
||
|
||
本仓库是可复用 AI 技能的目录集合。每个技能都是一个自包含的指导与元数据包，智能体可按需加载以执行特定任务。
|
||
|
||
## 你将在这里找到什么
|
||
|
||
- 涵盖 UI/UX、性能优化、内容创作和数据处理等领域的精选技能包。
|
||
- 统一的结构规范，使技能可被工具发现、加载和执行。
|
||
- 智能体可直接阅读的编译指南，无需额外上下文。
|
||
|
||
## 技能的组织方式
|
||
|
||
- `skills/<skill-id>/` — 每个技能独占一个文件夹（支持命名空间，如 `skills/anthropic/<skill-id>/`）
|
||
- 每个技能文件夹包含：
|
||
- `README.md` — 范围、目标用户和使用说明
|
||
- `SKILL.md` — 名称、描述和触发条件
|
||
- `AGENTS.md` — 为智能体编译的指南
|
||
- `metadata.json` — 版本和归属信息
|
||
- `rules/` — 源规则文件和模板
|
||
|
||
## 子智能体的组织方式
|
||
|
||
本仓库还在一个并列的顶层文件夹中存储子智能体定义：
|
||
|
||
- `subagents/` — 以带 YAML 前置元数据的 Markdown 文件存储的子智能体定义
|
||
|
||
子智能体是具有独立系统提示、工具访问权限和权限设置的专用助手。当任务匹配其描述时，Claude 会将任务委托给它们，从而有助于保留主上下文、强制执行约束并控制成本。它们常用于高频任务（如探索、测试输出或研究），以保持主对话的专注性。
|
||
|
||
关于子智能体的行为、作用域和配置，请参阅 Claude Code 文档：https://code.claude.com/docs/en/sub-agents
|
||
|
||
## 当前技能列表
|
||
|
||
`skills/` 中的当前技能文件夹：
|
||
|
||
- `action-creator`
|
||
- `add-admin-api-endpoint`
|
||
- `add-malli-schemas`
|
||
- `add-pattern`
|
||
- `add-uint-support`
|
||
- `agent-guardrails`
|
||
- `agent-identifier`
|
||
- `algorithmic-art`
|
||
- `algorithmic-art-modelscope`
|
||
- `analyzing-financial-statements`
|
||
- `app-review`
|
||
- `applying-brand-guidelines`
|
||
- `artifacts-builder`
|
||
- `ask-questions-if-underspecified`
|
||
- `at-dispatch-v2`
|
||
- `attack-tree-construction`
|
||
- `auth-implementation-patterns`
|
||
- `brand-guidelines`
|
||
- `cache-components`
|
||
- `campaign-planner`
|
||
- `candidate-evaluation`
|
||
- `canvas-design`
|
||
- `changelog-generator`
|
||
- `changeset-validation`
|
||
- `clojure-review`
|
||
- `clojure-write`
|
||
- `code-change-verification`
|
||
- `code-change-verification-js`
|
||
- `command-development`
|
||
- `command-name`
|
||
- `competitive-ads-extractor`
|
||
- `competitive-landscape`
|
||
- `concept-workflow`
|
||
- `configured-agent`
|
||
- `connect`
|
||
- `connect-apps`
|
||
- `connect-apps-plugin`
|
||
- `content-creator`
|
||
- `content-prd`
|
||
- `content-research-writer`
|
||
- `cookbook-audit`
|
||
- `create-database-migration`
|
||
- `create-plan`
|
||
- `create-pr`
|
||
- `create-unit-test-ask`
|
||
- `creating-financial-models`
|
||
- `creating-financial-models-modelscope`
|
||
- `data-storytelling`
|
||
- `developer-growth-analysis`
|
||
- `doc-coauthoring`
|
||
- `docs-review`
|
||
- `docs-sync`
|
||
- `docs-sync-js`
|
||
- `docs-write`
|
||
- `docstring`
|
||
- `document-skills`
|
||
- `docx`
|
||
- `domain-name-brainstormer`
|
||
- `electron-chromium-upgrade`
|
||
- `enter-services`
|
||
- `example-skill`
|
||
- `examples-auto-run`
|
||
- `examples-auto-run-js`
|
||
- `executive-briefing`
|
||
- `fact-check`
|
||
- `file-organizer`
|
||
- `final-release-review`
|
||
- `final-release-review-js`
|
||
- `find-bugs`
|
||
- `frontend-design`
|
||
- `g2-legend-expert`
|
||
- `g2-translation-guidelines`
|
||
- `g2-unit-testing-skills`
|
||
- `gemini-pr-creator`
|
||
- `gemini-skill-creator`
|
||
- `gerrit-storj`
|
||
- `gh-address-comments`
|
||
- `hook-development`
|
||
- `image-enhancer`
|
||
- `implementing-jsc-classes-cpp`
|
||
- `implementing-jsc-classes-zig`
|
||
- `infographic-creator`
|
||
- `infographic-item-creator`
|
||
- `infographic-structure-creator`
|
||
- `infographic-syntax-creator`
|
||
- `infographic-template-updater`
|
||
- `integration-tests`
|
||
- `internal-comms`
|
||
- `invoice-organizer`
|
||
- `issue-maker`
|
||
- `kimi-cli-help`
|
||
- `kpi-dashboard-design`
|
||
- `landing-page-optimizer`
|
||
- `langsmith-fetch`
|
||
- `lead-research-assistant`
|
||
- `listener-creator`
|
||
- `mcp-builder`
|
||
- `mcp-integration`
|
||
- `meeting-insights-analyzer`
|
||
- `model-debugging`
|
||
- `monitor-experiment`
|
||
- `notion-knowledge-capture`
|
||
- `notion-spec-to-implementation`
|
||
- `openai-knowledge`
|
||
- `openai-knowledge-js`
|
||
- `openai-skill-creator`
|
||
- `openai-skill-creator-system`
|
||
- `openai-skill-installer`
|
||
- `openai-skill-installer-system`
|
||
- `patterns`
|
||
- `pdf`
|
||
- `plugin-settings`
|
||
- `postmortem-writing`
|
||
- `pptx`
|
||
- `pr-draft-summary`
|
||
- `pr-draft-summary-js`
|
||
- `product-prd`
|
||
- `product-strategist`
|
||
- `r2-glacier-migration`
|
||
- `raffle-winner-picker`
|
||
- `react-best-practices`
|
||
- `readme-writing`
|
||
- `resource-curator`
|
||
- `rule-identifier`
|
||
- `rust-errors`
|
||
- `sast-configuration`
|
||
- `scanning-tools`
|
||
- `security-compliance`
|
||
- `senior-security`
|
||
- `seo-review`
|
||
- `single-or-array-pattern`
|
||
- `skill-creator`
|
||
- `skill-creator-kimi`
|
||
- `skill-development`
|
||
- `skill-development-official`
|
||
- `skill-share`
|
||
- `skill-writer`
|
||
- `slack-gif-creator`
|
||
- `social-media`
|
||
- `solidity-security`
|
||
- `stripe-best-practices`
|
||
- `styling`
|
||
- `svelte`
|
||
- `tailored-resume-generator`
|
||
- `template-skill`
|
||
- `tempo-codegen`
|
||
- `test-coverage-improver`
|
||
- `test-coverage-improver-js`
|
||
- `test-with-spanner`
|
||
- `test-writer`
|
||
- `theme-factory`
|
||
- `tier-management`
|
||
- `tldraw-skill-creator`
|
||
- `typescript`
|
||
- `typescript-review`
|
||
- `typescript-write`
|
||
- `vercel-react-best-practices`
|
||
- `video-downloader`
|
||
- `voting-status`
|
||
- `web-artifacts-builder`
|
||
- `webapp-testing`
|
||
- `workflow`
|
||
- `write-concept`
|
||
- `write-docs`
|
||
- `write-e2e-tests`
|
||
- `write-example`
|
||
- `write-issue`
|
||
- `write-pr`
|
||
- `write-unit-tests`
|
||
- `writing-bundler-tests`
|
||
- `writing-dev-server-tests`
|
||
- `xlsx`
|
||
- `zig-system-calls`
|
||
|
||
## 了解更多
|
||
|
||
打开任意技能文件夹，从其 `README.md` 开始，了解该技能的覆盖范围及使用方式。更多详情，请访问 [https://aiagentskills.xyz/](https://aiagentskills.xyz/)。 |