cancel_job
Cancel a queued or running indexing job. Use the 'list_jobs' tool to find the job ID.
LABSCO SUMMARY
Version-specific docs, indexed once and read locally.
The failure this addresses is an assistant confidently using an API that changed two releases ago. Because the index is keyed by version and the search resolves against it, the answer comes from the docs for the version in your lockfile — and the whole thing stays on your machine.
What it is
A personal documentation index. It fetches official docs from websites, GitHub, npm, PyPI, local folders and archives, stores them by library and version, and answers searches from that store — so an assistant reads the version in your project rather than whatever it remembers. Everything runs on your machine.
What you get
- A documentation set scraped into the index under a library name and version, with the crawl depth and scope under your control
- Search over the indexed docs, scoped to the version you name, with a lookup that resolves which indexed version best matches the one you have
- The libraries and versions already indexed, listed
- A stored set refreshed when the upstream docs move on, reusing the settings the original scrape ran with, and removed when you no longer want it
- Any single URL fetched and converted to Markdown on its own, without indexing anything
- Long scrapes handled as jobs — listed, inspected while they run, and cancelled
- A wide reach of formats: HTML and Markdown, PDF, Word, Excel, PowerPoint, OpenDocument, RTF, EPUB, Jupyter notebooks, archives that get unpacked, and source files across 90+ languages
- Markdown preferred at the source: the scraper probes for llms.txt and asks for Markdown by content negotiation, so a server that offers it skips the HTML-to-Markdown conversion entirely
- Hash-routed documentation sites handled on request, with the job upgraded to a real browser automatically because plain fetching cannot resolve client-side routes
Requirements
Node.js 22 or newer, run with npx from @arabold/docs-mcp-server, or from the ghcr.io container with volumes for its data and config. It is a long-running local server: the web UI for managing the index is at http://localhost:6280 and clients connect to the SSE endpoint on the same port. Nothing leaves your network, and no account is involved. An embedding model is optional but is what turns keyword matching into semantic search — set an OpenAI key in the environment, or point it at Ollama, Gemini or Azure instead.
Setup effort
One command — npx @arabold/docs-mcp-server@latest
USE CASES
01
Grounding an assistant in the version you actually depend on
Scrape the library's docs at that version once, then search them whenever a question about it comes up.
02
Answering a question about an API you half remember
Search the index rather than the open web, and read the passage from the official docs it returns.
03
Keeping the index current after an upgrade
Refresh the stored version with the settings it was scraped under, and drop the sets you no longer support.
04
Indexing something big without blocking
Start the scrape, watch the job while it runs, and cancel it if it is crawling somewhere you did not intend.
05
Reading one page without committing to a whole index
Fetch the URL as Markdown for a one-off question — nothing is stored and nothing is crawled.
06
Making internal docs searchable
Point a scrape at a local folder, a GitHub repository or an archive, and the same search covers material that was never on a docs site.
10 tools
Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.
fetch_url
Fetch a single URL and convert its content to Markdown. Use this tool to read the content of any web page.
find_version
Find the best matching version for a library. Use to identify available or closest versions.
get_job_info
Get details for a specific indexing job. Use the 'list_jobs' tool to find the job ID.
list_jobs
List all indexing jobs. Optionally filter by status.
list_libraries
List all indexed libraries.
refresh_version
Re-scrape a previously indexed library version, updating only changed pages.
remove_docs
Remove indexed documentation for a library version. Use only if explicitly instructed.
scrape_docs
Scrape and index documentation from a URL for a library. Use this tool to index a new library or a new version.
search_docs
Search up-to-date documentation for a library or package. Examples:
- {library: "react", query: "hooks lifecycle"} -> matches latest version of React
- {library: "react", version: "18.0.0", query: "hooks lifecycle"} -> matches React 18.0.0 or earlier
- {library: "typescript", version: "5.x", query: "ReturnType example"} -> any TypeScript 5.x.x version
- {library: "typescript", version: "5.2.x", query: "ReturnType example"} -> any TypeScript 5.2.x version
BEFORE YOU INSTALL
What to know first
Tools you get
10 tools when we connected.
Requires
npx on PATH
CONTEXT COST
Light Standing cost
Rated by the size of the 10 tool definitions that ride along on every turn, not by how many there are
INSTALLATION
Connect your client
Server setup
TERMINAL
npx @arabold/docs-mcp-server@latestMCP Clients
Claude Desktop, Windsurf, LM Studio, Roo Code, Trae, Gemini CLI, Amazon Q, JetBrains IDEs
JSON
{
"mcpServers": {
"docs-mcp-server": {
"command": "npx",
"args": [
"-y",
"@arabold/docs-mcp-server@latest"
]
}
}
}Cursor
Go to Settings → Cursor Settings → MCP
Click Add new MCP server
Set Type to stdio and Command to npx -y @arabold/docs-mcp-server@latest
Zed
ZED · JSON
{
"context_servers": {
"docs-mcp-server": {
"command": "npx",
"args": [
"-y",
"@arabold/docs-mcp-server@latest"
]
}
}
}Cline — remote
CLINE — REMOTE · JSON
{
"mcpServers": {
"docs-mcp-server": {
"url": "http://localhost:6280/mcp",
"type": "streamableHttp"
}
}
}Continue
CONTINUE · JSON
{
"mcpServers": [
{
"name": "docs-mcp-server",
"command": "npx",
"args": [
"-y",
"@arabold/docs-mcp-server@latest"
]
}
]
}Claude Code
CLAUDE CODE · TERMINAL
claude mcp add --scope user docs-mcp-server -- npx -y @arabold/docs-mcp-server@latestOpenCode
OPENCODE · JSON
{
"mcp": {
"docs-mcp-server": {
"type": "local",
"command": [
"npx",
"-y",
"@arabold/docs-mcp-server@latest"
],
"enabled": true
}
}
}GitHub Copilot
GITHUB COPILOT · JSON
{
"mcpServers": {
"docs-mcp-server": {
"type": "local",
"command": "npx",
"args": [
"-y",
"@arabold/docs-mcp-server@latest"
]
}
}
}Visual Studio
VISUAL STUDIO · JSON
{
"servers": {
"docs-mcp-server": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@arabold/docs-mcp-server@latest"
]
}
}
}Any MCP client — Docker
ANY MCP CLIENT — DOCKER · JSON
{
"mcpServers": {
"docs-mcp-server": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-v",
"docs-mcp-data:/data",
"ghcr.io/arabold/docs-mcp-server:latest"
]
}
}
}
