Skip to main content
Modern AI coding assistants work better when they can access your documentation directly. Rather than relying on outdated training data, tools like Claude, Cursor, and VS Code Copilot can query Dfns docs in real-time—giving you accurate answers about our APIs, authentication flows, and blockchain integrations.

Quick Setup by Tool

  1. Open Claude Settings → Connectors
  2. Click Add custom connector
  3. Enter name Dfns and URL https://docs.dfns.co/mcp
  4. When chatting, click the attachment icon and select Dfns to enable doc search
claude mcp add --transport http Dfns https://docs.dfns.co/mcp
  1. Open command palette: Cmd+Shift+P (Mac) or Ctrl+Shift+P (Windows)
  2. Search Open MCP settingsAdd custom MCP
  3. Add to mcp.json:
{
  "mcpServers": {
    "Dfns": {
      "url": "https://docs.dfns.co/mcp"
    }
  }
}
  1. Open command palette: Cmd+Shift+P (Mac) or Ctrl+Shift+P (Windows)
  2. Run MCP: Add Server and select Workspace (project-specific) or User (global)
  3. Add the following configuration:
{
  "servers": {
    "Dfns": {
      "type": "http",
      "url": "https://docs.dfns.co/mcp"
    }
  }
}
See VS Code MCP documentation for more options.

How It Works

Dfns documentation is automatically optimized for AI consumption through the Model Context Protocol (MCP)—an open standard that lets AI tools search and retrieve documentation content. When you connect an AI tool to our MCP server https://docs.dfns.co/mcp, it gains the ability to:
  • Search across all Dfns documentation
  • Retrieve relevant pages based on your questions
  • Access the latest content (not stale training data)
The AI decides when to search based on conversation context. Ask about creating wallets, signing transactions, or configuring policies, and it will automatically pull the relevant documentation.

Bulk Access for Custom Integrations

Building your own AI tooling or RAG pipeline? Access the full documentation programmatically:
ResourceDescription
/llms.txtIndex of all pages with descriptions
/llms-full.txtComplete documentation as a single file
Any page is also available as Markdown by adding .md to its URL (e.g., /api-reference.md).