agent-fecfile

FEC Filing Plugin for Claude Code

A Claude Code plugin for analyzing Federal Election Commission (FEC) campaign finance filings. Includes an Agent Skill and an MCP server for API access.

This plugin enables AI agents to fetch, parse, and analyze FEC filings directly within agent sessions. Parsing and filtering happen outside the model context, allowing agents to programmatically reduce large filings before analysis, saving tokens and enabling efficient queries against filings of any size.

The plugin includes detailed field mappings for common form types and schedules, helping agents accurately interpret campaign finance data like contributions, disbursements, and committee information.

Features

• Fetch and analyze FEC filings by filing ID (Agent Skill)
• Search for committees and filings via the FEC API (MCP server)
• Support for major form types (F1, F2, F3, F99)
• Detailed field mappings for contributions, disbursements, and schedules
• Auto-installing dependencies via uv

Requirements

• Claude Code CLI or another MCP-compatible runtime (e.g., Codex CLI)
• uv (for running Python scripts)
• Python 3.9+
• An FEC API key (for committee/filing search)

Installation

Claude Code Plugin (Recommended)

Install via the Claude Code plugin system:

# Add the marketplace
claude plugin marketplace add hodgesmr/agent-fecfile

# Install the plugin
claude plugin install fecfile@agent-fecfile

You may need to restart your Claude Code session to properly load the Agent Skill and MCP Server.

When installed:

• The Agent Skill (fecfile) is automatically available
• The MCP server starts automatically, providing search_committees and get_filings tools

Updating:

# Update the marketplace
claude plugin marketplace update agent-fecfile

# Update the plugin
claude plugin update fecfile@agent-fecfile

Other Compatible Runtimes (Codex, etc.)

For agent runtimes that support Agent Skills and MCP but not Claude Code plugins:

1. Clone the repository:

git clone --branch latest git@github.com:hodgesmr/agent-fecfile.git ~/agent-fecfile

2. Install the Agent Skill by symlinking to your runtime's skills directory:

# Codex CLI Global install
ln -sfn ~/agent-fecfile/skills/fecfile ~/.codex/skills/fecfile

3. Configure the MCP server using your runtime's MCP configuration:

# Codex CLI
codex mcp add fec-api -- uv run ~/agent-fecfile/mcp-server/server.py

[!IMPORTANT] The MCP server loads the FEC API key from the system keyring on first tool use. You should expect to see a system prompt to authorize Python's access to the key the first time you use search_committees or get_filings.

Updating:

cd ~/agent-fecfile && git fetch --tags --force && git checkout latest

Usage

Once installed, ask your agent to analyze FEC filings.

[!TIP] For best results, use the most capable models available in your agent runtime (e.g., Claude Opus 4.5 or comparable frontier models). More capable models produce better analysis and more accurate field interpretations.

Basic Usage (With A Filing ID)

If you already have an FEC filing ID, you can work with it directly, without needing an API key:

❯ What are the largest expenditures in filing 1896830?

⏺ FEC Filing 1896830 - Largest Expenditures

  | #   | Recipient                  |    Amount | Date       | Purpose          |
  |-----|----------------------------|-----------|------------|------------------|
  | 1   | Trump National Doral Hotel | $1,013.61 | 2025-05-19 | RNC training     |
  | 2   | Uber.com                   |    $19.77

... [View full README on GitHub](https://github.com/hodgesmr/agent-fecfile#readme)