Is this the official DeepWiki MCP server?
No. This listing is `regenrek/deepwiki-mcp`, an unofficial community project. The official DeepWiki MCP server is run by Devin and documented at `https://docs.devin.ai/work-with-devin/deepwiki-mcp`.
@regenrek
Basic information
📖 MCP server for fetch deepwiki.com and get latest knowledge in Cursor and other Code Editors
Add this server to your MCP-compatible client using the configuration below.
{
"mcpServers": {
"mcp-deepwiki": {
"command": "node",
"args": [
"./bin/cli.mjs"
]
}
}
}Tool list not available yet
This listing does not currently have a tool list to display. This does not mean the server has no tools; check its documentation for supported capabilities.
We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.
This DeepWiki MCP server is an unofficial, community-built tool (mcp-deepwiki by regenrek) that crawled deepwiki.com pages and returned them to your editor as Markdown. Its README says it no longer works because DeepWiki blocked scraping, and recommends the official DeepWiki MCP server instead.
DeepWiki generates documentation wikis for GitHub repositories. The official DeepWiki MCP server, run by Devin, is a free remote server at https://mcp.deepwiki.com/mcp that needs no authentication for public repositories. If you searched for "deepwiki mcp" to give Cursor or Claude context about a repository, the official server is the one to install. This page covers both, clearly labeled.
No install, API key or account is needed for public repositories. Use the Streamable HTTP endpoint https://mcp.deepwiki.com/mcp. The SSE endpoint https://mcp.deepwiki.com/sse still exists but is being deprecated.
claude mcp add -s user -t http deepwiki https://mcp.deepwiki.com/mcp
Add to .cursor/mcp.json (the same url format works in other clients that use it):
{
"mcpServers": {
"deepwiki": {
"url": "https://mcp.deepwiki.com/mcp"
}
}
}
Add https://mcp.deepwiki.com/mcp as a remote server through Claude's custom connector settings. Devin's docs link to Anthropic's guide for connecting remote MCP servers rather than giving a config file.
For private repositories, Devin's docs point to the separate Devin MCP server, which requires a Devin account and API key.
The project's original Cursor setup in .cursor/mcp.json was:
{
"mcpServers": {
"mcp-deepwiki": {
"command": "npx",
"args": ["-y", "mcp-deepwiki@latest"]
}
}
}
It can also be built from source with git clone https://github.com/regenrek/deepwiki-mcp.git, npm install and npm run build. Expect fetch errors: according to the README notice, DeepWiki has cut off the scraping this server relies on.
These are the community server's designed features, as listed in its README:
--http) transportThe official server, by contrast, reads DeepWiki's generated wiki directly and can answer questions about a repository.
The community server registers one tool, deepwiki_fetch. It takes url (required; the starting deepwiki.com URL), mode (aggregate for one Markdown document, the default, or pages for structured pages) and maxDepth (default 10). It returns ok, partial or error responses with page counts and timing.
The official DeepWiki MCP server exposes three different tools:
read_wiki_structure: lists the documentation topics for a GitHub repositoryread_wiki_contents: returns the documentation about a repositoryask_question: answers a question about a repository with an AI-generated, context-grounded responseWith the official server, which is where these prompts work today:
The official DeepWiki MCP server is the direct replacement for this listing: it is maintained by DeepWiki's operator, needs no scraping and adds ask_question. For library documentation from package sources rather than DeepWiki, Context7 MCP serves version-specific docs to coding agents. To read a project's actual source, the GitHub MCP Server gives agents access to repositories, issues and pull requests.
deepwiki_fetch returns errors or empty results: this is the known failure described in the README notice. Switch to the official server at https://mcp.deepwiki.com/mcp.DOMAIN_NOT_ALLOWED error: the community server only accepts deepwiki.com URLs.DEEPWIKI_REQUEST_TIMEOUT and DEEPWIKI_MAX_CONCURRENCY, for example DEEPWIKI_REQUEST_TIMEOUT=60000 DEEPWIKI_MAX_CONCURRENCY=10 npx mcp-deepwiki.EACCES permission denied: make the binary executable with chmod +x ./node_modules/.bin/mcp-deepwiki.serverUrl field; using the wrong field name causes the server to be silently ignored.No. This listing is `regenrek/deepwiki-mcp`, an unofficial community project. The official DeepWiki MCP server is run by Devin and documented at `https://docs.devin.ai/work-with-devin/deepwiki-mcp`.
Its own README says it is currently not working, because DeepWiki cut off the ability to scrape the site. The author recommends the official server for the time being.
Both are free to use. The community project is MIT-licensed, and the official server is described by Devin as a free, remote, no-authentication service for public repositories.
Not for public repositories on the official server. Private repositories require the separate Devin MCP server with a Devin account and API key. The community server never needed a key.
Yes. Devin's docs give a Claude Code command (`claude mcp add -s user -t http deepwiki https://mcp.deepwiki.com/mcp`) and a Cursor config using the `url` field.
`https://mcp.deepwiki.com/mcp` uses Streamable HTTP and is recommended. `https://mcp.deepwiki.com/sse` is the legacy SSE endpoint and is being deprecated.
Not through the free public server. Devin's docs direct private-repository use to the Devin MCP server, which needs a Devin account.

ArtiBot Cloud MCP (Zhijiangyun 智匠云) — cloud MCP server for robot manipulation data & GPU workloads. 15 tools: robot fleet management, data collection & video recording, post-processing pipelines, and benchmark evaluation
Gives AI agents public URLs (tunnels) for localhost, live HTTP traffic inspection, snapshot publishing, and access control. Part of LocalCan, the ngrok alternative for Mac, Windows and Linux. Free plan.

Agents make claims. Reelier writes receipts — record an agent's tool-call workflow once, replay it deterministically at 0 tokens, and diff runs to catch drift.
Set up and manage the Eddyter AI rich text editor by talking to your AI agent — for Claude, Cursor, Codex and other MCP tools. A Model Context Protocol server that lets AI tools add Eddyter to a React, Next.js, Vite or R

Official Porkbun MCP server, maintained by Porkbun. Check domain availability and pricing, register, renew and transfer domains, and manage DNS records, nameservers, URL forwarding, glue records, DNSSEC and domain contac
Comments