sec-edgar

The SEC EDGAR MCP server provides access to SEC EDGAR filings and related financial data by connecting AI assistants to company filings, financial statements, and insider trading data with precise numeric information. It exposes tools under categories such as Company (CIK lookup, company info, company facts), Filings (retrieval of 10-K, 10-Q, 8-K and section extraction), Financials (balance sheet, income statement, cash flow with XBRL-parsed data), and Insider Trading (Form 3/4/5 transactions). All responses include SEC filing URLs for verification, enabling traceability of the source documents.

To run the server, start it via Docker as shown in the Quick Start example. Once running, you can query the available tools through your MCP client, and request specific operations like looking up a company by CIK, retrieving a 10-K filing, or extracting a table from a financial statement. For HTTP-based integrations, you can use the streamable-http transport option to send and receive data over HTTP instead of standard I/O, as documented in the Quick Start section. The server is designed to be integrated with AI assistants and other MCP-enabled clients, enabling precise access to SEC filings within conversational or task-oriented workflows.

Prerequisites:

• Docker installed on your machine or environment where you plan to run MCP servers
• Optional: Python environment if you prefer non-Docker installations (see documentation link in README for details)
• Access to the SEC EDGAR MCP package and any required tokens or environment configurations per your setup

Installation steps (Docker):

1. Ensure Docker is running on your system.
2. Start the MCP server using the docker command from the Quick Start:

# Run the SEC EDGAR MCP server via Docker
docker run -i --rm \
  -e SEC_EDGAR_USER_AGENT="Your Name (your@email.com)" \
  stefanoamorelli/sec-edgar-mcp:latest

This command launches the MCP server container and enables JSON-RPC communication. The -i flag is required for MCP's JSON-RPC communication as noted in the README.

Alternative installation methods (see official docs):

• pip (Python): Install the sec-edgar-mcp package and run the Python server module as described in the project docs.
• conda (Python): Create a conda environment and install the package per the docs, then run the server.
• uv (Python/uvicorn based): Install via uvx and run the server module as documented.

For exact commands and environment variable details, refer to the project Quick Start and Documentation pages linked in the README.

Notes and tips:

• The SEC_EDGAR_USER_AGENT environment variable is required to identify the client in SEC EDGAR requests. Replace with your name and contact email.
• The Quick Start example uses docker run with --rm to ensure a clean ephemeral container; remove --rm if you want to keep the container running after stop for debugging.
• If you plan to expose the server over HTTP transport (e.g., for Dify integration), follow the HTTP Transport guidance in the docs to use streamable-http instead of stdio.
• Always verify the SEC URLs returned from queries against the official EDGAR filings for traceability.
• If you run evaluations or tests, you may need to install the evals suite as shown in the Eval section of the README (Promptfoo-based tests).