Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.frontic.com/llms.txt

Use this file to discover all available pages before exploring further.

The Frontic Docs MCP server is a read-only endpoint that exposes the Frontic documentation to any MCP-compatible AI agent. Connect it alongside Frontic Tools and your editor’s agent can answer “how does X work?” questions by querying the docs directly — no leaving the editor, no tab-switching to read. The docs MCP needs no authentication. It’s a public, read-only resource — same one this site is built from.

​
What it’s good for

Answering how-to questions inline

Your agent looks up the right documentation page when you ask something like “how do I configure a domain?” — pulling the answer into the conversation rather than guessing.

Grounding generated code

When the agent writes against the Frontic SDK or API, it can cross-reference the docs to make sure it’s using the right method shape, the right parameters, the right defaults.

Onboarding teammates

A new dev’s editor agent can answer Frontic-specific questions immediately, instead of asking them to dig through the docs site while learning the codebase.

Cross-checking against changes

The docs MCP always serves the current docs site, so an agent’s answer about a feature is as fresh as the latest documentation publish.

​
Setup

The fastest path is the CLI — pick your package manager: 
frontic mcp init --client cursor
Swap cursor for claude, windsurf, vscode, or codex. The CLI prompts you to pick what to install — choose Docs only (or Tools & Docs if you also want Frontic Tools). See the Frontic CLI reference for full options. If you’d rather edit the config yourself, drop this entry into your client’s MCP config file: 
{
  "mcpServers": {
    "frontic-docs": {
      "type": "http",
      "url": "https://docs.frontic.com/mcp"
    }
  }
}
No authentication is required — the Frontic Docs MCP is public and read-only.

​
Client-specific configuration

Add to .mcp.json at your project root — Claude Code reads project MCP config from this file:
{
  "mcpServers": {
    "frontic-docs": {
      "type": "http",
      "url": "https://docs.frontic.com/mcp"
    }
  }
}
Edit the Claude Desktop config file:
  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows: %APPDATA%\Claude\claude_desktop_config.json
{
  "mcpServers": {
    "frontic-docs": {
      "type": "http",
      "url": "https://docs.frontic.com/mcp"
    }
  }
}
Add to .cursor/mcp.json in your project root, or Cursor’s global settings:
{
  "mcpServers": {
    "frontic-docs": {
      "type": "http",
      "url": "https://docs.frontic.com/mcp"
    }
  }
}

​
Example prompts

Once connected, ask questions you’d otherwise look up by hand:

Ask Frontic a concept question without leaving the editor

Look up an SDK response shape from the docs

Get a step-by-step answer for how to set up a feature

Have the docs MCP pull a reference catalogue for you

The agent calls the docs MCP, retrieves the relevant section, and answers from it — usually citing the page so you can follow up if you need more depth.

​
Pairing with Frontic Tools

The two MCP servers compose naturally: Frontic Tools acts on your project, Frontic Docs explains the platform. With both connected, an agent can read the docs to learn what it should be doing, then use the project tools to do it. That’s the recommended setup for any IDE-side Frontic workflow — install both via frontic mcp init and pick Tools & Docs at the prompt.

Frontic Tools

The MCP server that acts on your Frontic project — full tool catalogue.

Frontic CLI

The frontic mcp init command that wires both MCP servers into your client.

MCP Tools

Editor-facing walkthrough for setting up Frontic MCP in Claude Code, Cursor, and friends.