MCP / AI

The AskEdgar API documentation includes a built-in MCP (Model Context Protocol) server. Once connected, AI tools like Cursor, Claude Desktop, Windsurf, and VS Code Copilot can search our docs, understand our endpoints, and make authenticated API calls on your behalf — all from within your editor.

What you'll need

• An AskEdgar account
• An API key from the API Platform
• An MCP-compatible AI tool (Cursor, Claude Desktop, VS Code, Windsurf, etc.)

Connect to the MCP server

Add this to your MCP client configuration:

{
  "mcpServers": {
    "AskEdgar": {
      "url": "https://askedgar.readme.io/mcp",
      "headers": {
        "authorization": "Bearer your_api_key"
      }
    }
  }
}

⚠️

The authorization header is forwarded by the MCP server to the AskEdgar API when executing requests. Replace your_api_key with your actual key from the API Platform.

Where to add this config

The config file location depends on your AI tool:

Cursor Settings → Features → MCP Servers → Add Server, or edit ~/.cursor/mcp.json directly.

Claude Desktop Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows).

VS Code (Copilot) Add to your workspace .vscode/mcp.json or user settings.

Windsurf Edit ~/.windsurf/mcp_config.json.

For clients that don't support direct URL connections, use the mcp-remote bridge:

{
  "mcpServers": {
    "AskEdgar": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://askedgar.readme.io/mcp",
        "--header",
        "Authorization:${ASKEDGAR_API_KEY}"
      ],
      "env": {
        "ASKEDGAR_API_KEY": "Bearer your_api_key"
      }
    }
  }
}

What you can do

Once connected, your AI assistant gets access to two categories of tools:

API tools — generated directly from our OpenAPI spec. Your assistant can look up any endpoint's parameters, schemas, and response formats, then execute live API calls.

Documentation tools — search and read our full documentation, including guides, error handling, and endpoint descriptions.

Example prompts

Here are a few things you can ask your AI assistant once connected:

• "What endpoints does AskEdgar have for dilution data?"
• "Get the dilution rating for ticker MULN"
• "Show me the float and outstanding shares for FFIE"
• "What offerings has NVAX had in the last 6 months?"
• "Search the AskEdgar docs for how pagination works"

Your assistant will figure out which endpoints to call, what parameters to use, and how to interpret the response — all from the OpenAPI spec.

Authentication and credits

MCP requests that execute API calls consume credits from your account, just like direct API calls. The cost estimate endpoint is free and can be called through MCP to check costs before running a query.

Your API key is passed via the authorization header in your MCP config and forwarded to the AskEdgar API on every executed request. If your key is invalid or you're out of credits, you'll see the same error responses as with direct API usage.

🔑

Your API key is long-lived and doesn't expire automatically. If you need to rotate it, generate a new one in the API Platform and update your MCP config.

Troubleshooting

"Invalid or missing API key" errors Make sure your authorization header includes the Bearer prefix: "Bearer your_api_key", not just the key itself.

AI can't find endpoints Restart your AI tool after adding the MCP config. Some clients cache the server connection on startup.

"Insufficient credits" errors Add credits via the API Platform. You can estimate costs beforehand with the pricing calculator.

Next steps