Query any public GitHub repository to get AI-powered documentation and source code answers without needing an API key.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install deepwiki-mcp
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install deepwiki-mcp using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is DeepWiki MCP?
The DeepWiki MCP skill is a specialized tool designed to bridge the gap between AI agents and public GitHub codebases. By leveraging the Model Context Protocol (MCP), these Openclaw Skills allow users to ask complex, natural-language questions about a repository's internal architecture, configuration, and source code logic. It serves as an AI-powered documentation layer that grounded answers in the actual files of a project, providing much deeper insights than standard search tools.
This skill is particularly useful for developers exploring new frameworks or contributors trying to understand specific implementations within a codebase. Because it requires no API keys or authentication for public repositories, it offers a frictionless way to integrate deep repository knowledge into any agent-based workflow. Whether you are investigating the OpenClaw source or a major library like React, this skill provides the context needed to work efficiently.
DeepWiki MCP Use Cases
- Understanding complex logic by asking how specific functions or features work within a codebase.
- Onboarding to a new project by querying the repository architecture and directory structure.
- Searching for specific configuration patterns or internal documentation across a public repo.
- Automating technical research for AI agents that need grounded facts from source code.
How DeepWiki MCP Works
- The user provides a natural-language question and an optional repository name (defaulting to OpenClaw).
- The skill invokes a bash helper script or a direct curl command to communicate with the DeepWiki MCP endpoint.
- The request is processed by DeepWiki's AI engine, which analyzes the repository's indexed code and documentation.
- The endpoint returns a JSON-RPC response containing an AI-generated answer with specific file references.
- The Openclaw Skills agent parses the response and presents a substantive, grounded answer to the user.
DeepWiki MCP Setup
To utilize these Openclaw Skills, your environment must have bash, curl, and python3 available for script execution. No persistent configuration or API keys are necessary. You can use the included helper script:
# Ask a question about the OpenClaw repository
./scripts/deepwiki.sh ask "How does session compaction work?"
# Query a different public repository
./scripts/deepwiki.sh ask facebook/react "How does concurrent mode work?"
DeepWiki MCP Data Schema & Taxonomy
The skill organizes data and interactions through specific parameters and structured outputs:
| Field | Description |
|---|---|
| Input: Repository | The owner/repo string for any public GitHub project. |
| Input: Action | Supports ask (Q&A), topics (metadata), or docs (full content). |
| Output: Text | AI-grounded answer containing logic explanations and code paths. |
| Output: References | Citations of specific files and lines of code analyzed by the AI. |
DeepWiki MCP Advanced Features
- Dynamic indexing that triggers a fresh crawl when a new public repository is queried for the first time.
- Fallback direct curl functionality to maintain operations even if the primary helper script is missing.
- SSE (Server-Sent Events) stream processing for handling high-latency AI generation tasks.
- Specialized tools for reading entire wiki structures or fetching full documentation sets for offline processing.
SKILL.md
Loading
Related Openclaw Skills
Bird Twitter Skill
A high-speed CLI wrapper for Twitter/X that enables developers and AI agents to interact with the platform using GraphQL APIs.
Local AI Accounting Assistant
A private, local-first AI accounting tool for tracking finances through natural language conversation.
ClawSentinel
A local security auditor designed to detect ClawHavoc malware and MCP backdoors in AI agent skills before installation.
Remix API Key Auth
A specialized skill for configuring and verifying bearer API key authentication to enable secure Remix agent publishing workflows.
git-repo-to-book
A sophisticated multi-agent orchestration skill that automates the research, writing, and review of full-length technical books directly from code repositories.
Incident Response
A structured, 7-phase incident response framework for diagnosing and fixing OpenClaw system regressions and configuration failures.
Featured*
