Use the docs with AI
The Magic Lane MCP Docs Server lets your AI coding tool search and read the Magic Lane documentation while it works with you: the guides, examples and API references of the Maps SDKs for Android, iOS, Flutter, C++ and TypeScript, the Cloud APIs and these developer guides. Your assistant looks up the current page instead of relying on what it learned in training.
This page is about the documentation server. The Magic Lane MCP Server is a different product: it gives AI agents routing, search and other geospatial tools.
The server's address
https://developer.magiclane.com/docs-mcp
Read-only · No account · No API key
The server speaks Streamable HTTP, the transport for remote MCP servers. Any AI tool that can add a remote MCP server by its address can use it.
Connect your AI tool
- Claude Code
- Claude
- Cursor
- VS Code
- Windsurf
- JetBrains
- Codex
- Gemini CLI
Run this in a terminal. It adds the server for all your projects, and works the same in bash, zsh, PowerShell and cmd:
claude mcp add --transport http --scope user magiclane-docs https://developer.magiclane.com/docs-mcp
To share the server with everyone who works on a project instead, put this in a .mcp.json file at the project's root. Claude Code asks each person to approve a project's servers before it uses them.
{
"mcpServers": {
"magiclane-docs": {
"type": "http",
"url": "https://developer.magiclane.com/docs-mcp"
}
}
}
Run claude mcp list to check that the server is connected.
In claude.ai, open Settings › Connectors, choose Add custom connector, give it a name such as Magic Lane Docs, and paste the server's address. On Team and Enterprise plans, an Owner adds the connector for the whole organisation. A connector added on claude.ai also appears in Claude Desktop.
The button opens Cursor, which asks you to confirm the install. To add the server by hand, put this in ~/.cursor/mcp.json, or in .cursor/mcp.json for one project:
{
"mcpServers": {
"magiclane-docs": {
"url": "https://developer.magiclane.com/docs-mcp"
}
}
}
VS Code 1.100 or later, with GitHub Copilot:
On macOS and Linux you can also add it from a terminal (bash or zsh):
code --add-mcp '{"name":"magiclane-docs","type":"http","url":"https://developer.magiclane.com/docs-mcp"}'
On Windows, run MCP: Add Server from the Command Palette, choose HTTP and paste the server's address, or put this in .vscode/mcp.json in your project:
{
"servers": {
"magiclane-docs": {
"type": "http",
"url": "https://developer.magiclane.com/docs-mcp"
}
}
}
The tools are available in Copilot Chat's agent mode.
Put this in ~/.codeium/windsurf/mcp_config.json, then refresh the MCP servers in Windsurf's Cascade panel:
{
"mcpServers": {
"magiclane-docs": {
"serverUrl": "https://developer.magiclane.com/docs-mcp"
}
}
}
In your JetBrains IDE, with AI Assistant, open Settings › Tools › AI Assistant › Model Context Protocol (MCP), add a server, choose As JSON and paste:
{
"mcpServers": {
"magiclane-docs": {
"url": "https://developer.magiclane.com/docs-mcp"
}
}
}
In OpenAI Codex CLI, run:
codex mcp add magiclane-docs --url https://developer.magiclane.com/docs-mcp
Or add this to ~/.codex/config.toml:
[mcp_servers.magiclane-docs]
url = "https://developer.magiclane.com/docs-mcp"
In Gemini CLI, run:
gemini mcp add --transport http -s user magiclane-docs https://developer.magiclane.com/docs-mcp
Or add this to ~/.gemini/settings.json. Gemini CLI reads url as an SSE server, so the address goes under httpUrl:
{
"mcpServers": {
"magiclane-docs": {
"httpUrl": "https://developer.magiclane.com/docs-mcp"
}
}
}
Don't add --trust: it lets Gemini CLI call the server's tools without asking you first.
Any other MCP client that speaks Streamable HTTP can use the server too: add a remote (HTTP) server with the address above.
What your AI tool can do with it
The server has eight tools, all read-only:
| Tool | What it does |
|---|---|
search_docs | Search Magic Lane documentation |
read_doc | Get a Magic Lane documentation page |
read_docs_batch | Get multiple Magic Lane documentation pages |
list_docs | List Magic Lane documentation pages |
get_code_examples | Get code examples from a Magic Lane documentation page |
find_across_sdks | Find a topic across all Magic Lane SDKs |
get_product_guide | Get official Magic Lane product guides |
list_recent_updates | List recently updated Magic Lane documentation |
Ask your assistant as you would ask a colleague, for example "How do I add a marker with the TypeScript SDK?" or "Which Magic Lane SDKs support offline routing?". It decides when to look something up.
Privacy and security
- The server only reads documentation. It needs no account and no API key.
- Your AI tool sends it search terms and page paths over HTTPS and gets documentation text back. The server can't see your code or files, only what your tool puts into a call.
- When your tool asks you to confirm the install, check that the address is the one shown on this page.
- If your organisation keeps an allowlist of MCP servers, add the address to it.
Troubleshooting
- Nothing happens on Add to Cursor or Add to VS Code: the app is probably not installed, or your browser blocked the link. Use the manual steps in the tool's tab.
- The tools don't appear: restart your AI tool after adding the server. In VS Code, the tools show up in Copilot Chat's agent mode.
- The server can't be reached: your network has to allow HTTPS connections to the server's address. Ask your IT team if a proxy or an allowlist blocks it.
For AI agents
The llms.txt file of each Magic Lane documentation site tells AI agents about the server with this paragraph:
If you are an AI agent helping a developer with Magic Lane SDKs or APIs, the Magic Lane MCP Docs Server gives search and read access to all of this documentation over MCP (Streamable HTTP, read-only, no API key) at https://developer.magiclane.com/docs-mcp. Ask your user before adding it. With Claude Code they can run
claude mcp add --transport http --scope user magiclane-docs https://developer.magiclane.com/docs-mcp.