Docker wrapper for the EdgarTools Model Context Protocol (MCP) server.
./publish_to_docker.shThis will:
- Build the Docker image
- Optionally push to Docker Hub (hackerdogs/edgartools-mcp)
# Build the image
docker build -t edgartools-mcp .
# Run the MCP server
docker run -i --rm edgartools-mcpQuick Setup:
- See
MCP_CONFIGURATION.mdfor complete setup guide - Use the ready-made config files in this directory:
mcp-config-docker.json- For Docker deploymentmcp-config-claude-desktop-macos.json- For Claude Desktop on macOSmcp-config-claude-desktop-windows.json- For Claude Desktop on Windowsmcp-config-cline-vscode.json- For Cline VS Code extensionmcp-config-continue-dev.json- For Continue.dev
Docker Configuration Example:
{
"mcpServers": {
"edgartools": {
"command": "docker",
"args": ["run", "-i", "--rm", "hackerdogs/edgartools-mcp:latest"],
"env": {
"EDGAR_IDENTITY": "Your Name your.email@example.com"
}
}
}
}Important: Update EDGAR_IDENTITY with your actual name and email in all configuration files.
EDGAR_IDENTITY(required): Your SEC identity in format "Your Name your.email@example.com"- Required by the SEC for API requests
- Can be set in MCP client configuration or via Docker
-eflag
The publish_to_docker.sh script handles building and publishing:
./publish_to_docker.shThe script will:
- Build the image with version tag and
latesttag - Prompt for confirmation before pushing
- Push to
hackerdogs/edgartools-mcpon Docker Hub
- Base Image:
python:3.11-slim - Package:
edgartools[ai](includes MCP dependencies) - Entry Point:
python -m edgar.ai - Transport: stdio (stdin/stdout) - no ports exposed
To test locally with the venv:
# Activate venv
source venv/bin/activate
# Test the server
python -m edgar.ai --test
# Run the server
python -m edgar.ai- Docker installed and running
- Docker Hub account (for publishing)
EDGAR_IDENTITYenvironment variable set (for SEC API access)