← 返回 Skills 市场
319
总下载
0
收藏
1
当前安装
1
版本数
在 OpenClaw 中安装
/install lh-deepwiki
功能描述
Query DeepWiki MCP service via Streamable HTTP to get GitHub repository documentation, wiki structure, and AI-powered Q&A.
使用说明 (SKILL.md)
lh-deepwiki
这是由龙虾哥（liuhedev）维护的 DeepWiki MCP 技能。相比原版，本版本修复了 API 协议变动导致的问题，改用新的 /mcp 端点（Streamable HTTP），并优化了参数传递和超时处理。
主要变更
- 协议升级: 从过时的 SSE 切换到新的 Streamable HTTP 端点。
- 参数修复: 将旧版的
repo参数统一更新为repoName。 - 超时优化: 默认超时时间增加至 120s，适配复杂问答场景。
- 开源维护: 托管于
lh-openclaw-kit，支持自主迭代。
命令说明
AI 问答 (Ask Question)
对 GitHub 公共仓库进行基于文档的 AI 问答。
node ./scripts/deepwiki.js ask \x3Cowner/repo> "你的问题"
读取 Wiki 结构 (Read Wiki Structure)
获取指定仓库的文档目录结构。
node ./scripts/deepwiki.js structure \x3Cowner/repo>
读取 Wiki 内容 (Read Wiki Contents)
查看指定路径的具体文档内容。
node ./scripts/deepwiki.js contents \x3Cowner/repo> \x3Cpath>
使用示例
问问 OpenClaw 是什么:
node ./scripts/deepwiki.js ask openclaw/openclaw "What is OpenClaw?"
获取 React 文档结构:
node ./scripts/deepwiki.js structure facebook/react
注意事项
- 基础服务地址:
https://mcp.deepwiki.com/mcp - 仅支持公开 GitHub 仓库。
- 无需 API Key 或身份验证。
安全使用建议
This skill appears to do what it claims: it POSTs JSON-RPC requests to https://mcp.deepwiki.com/mcp to fetch documentation/answers about public GitHub repos. Before installing, be aware: (1) you need Node.js to run the provided script (SKILL.md uses 'node' but the registry metadata omitted this requirement); (2) queries (repo names, paths, and questions) are sent to an external service — do not use it for sensitive or private data; (3) there is no API key or auth, so trustworthiness and privacy depend on the DeepWiki endpoint. If you want extra caution, review the script locally and run it in a network-isolated environment or verify the endpoint's trustworthiness before granting network access.
功能分析
Type: OpenClaw Skill
Name: lh-deepwiki
Version: 1.0.0
The skill bundle provides a client for the DeepWiki MCP service. The `SKILL.md` clearly outlines the purpose and commands, which involve making HTTPS requests to `https://mcp.deepwiki.com/mcp`. The `scripts/deepwiki.js` Node.js script handles user input by parsing command-line arguments and embedding them safely as string values within JSON payloads for HTTPS POST requests. There is no evidence of shell injection vulnerabilities, data exfiltration, unauthorized file system access, or malicious prompt injection attempts against the OpenClaw agent. The code's behavior is consistent with its stated purpose of querying a public GitHub documentation service.
能力评估
Purpose & Capability
The skill's name/description match the included script: it calls a DeepWiki MCP endpoint to read public GitHub repo docs and answer questions. Minor inconsistency: SKILL.md examples run 'node ./scripts/deepwiki.js' but the registry 'required binaries' section lists none — Node.js should be declared as a required runtime.
Instruction Scope
SKILL.md instructs running the provided script to call the DeepWiki MCP server for ask/structure/contents. The instructions do not request reading local files, other env vars, or unrelated system data. The script only sends the repo name, path, and question to the DeepWiki endpoint and prints results.
Install Mechanism
No install spec or remote downloads are present; the skill is instruction-only plus a small local script. Nothing is written to disk by an installer during install.
Credentials
The skill requires no credentials or secrets and the script does not read environment variables or config paths. It only makes outbound HTTPS requests to https://mcp.deepwiki.com/mcp, which is consistent with the claimed functionality.
Persistence & Privilege
always is false and the skill does not request persistent system-level privileges or modify other skills. It operates at runtime and can be invoked by the user or autonomously per platform defaults.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install lh-deepwiki - 安装完成后，直接呼叫该 Skill 的名称或使用
/lh-deepwiki触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.0
Initial release: Query DeepWiki MCP for GitHub repo docs
元数据
常见问题
LH DeepWiki 是什么？
Query DeepWiki MCP service via Streamable HTTP to get GitHub repository documentation, wiki structure, and AI-powered Q&A. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 319 次。
如何安装 LH DeepWiki？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install lh-deepwiki」即可一键安装，无需额外配置。
LH DeepWiki 是免费的吗？
是的，LH DeepWiki 完全免费（开源免费），可自由下载、安装和使用。
LH DeepWiki 支持哪些平台？
LH DeepWiki 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 LH DeepWiki？
由 liuhedev（@liuhedev）开发并维护，当前版本 v1.0.0。
推荐 Skills