About edgartools
edgartools started in November 2022 as a Python module to read SEC filings the way I wished I could. It does about a million downloads a month now, and over time it's grown into three things that work together: an open-source library, this blog, and a real-time platform at edgar.tools.
Here's the map.
The library — edgartools on PyPI
The Python library is the foundation. pip install edgartools, point it at a filing, get back well-typed data: 8-K items, XBRL financials, Form 3/4/5 ownership, 13F holdings, 424B prospectus details, Form ADV. The same API works on one filing or fifty years of them.
MIT-licensed. Free forever.
This blog — edgartools.io
This is where the thinking lives. Bug stories, design decisions, deep dives on weird corners of EDGAR, posts on AI tooling, the occasional Wall Street tangent. Subscribers get new posts about twice a month.
If you arrived here from Google searching for "how do I parse XBRL with Python" or "how to read Form 4," you're in the right place.
The platform — edgar.tools
The library is great for batch jobs and notebooks. But once you're running real workflows — watching for an 8-K from a portfolio company, alerting on a Form 4, feeding filings into an AI agent — you need something that's always on. That's edgar.tools.
edgar.tools turns SEC filings into a real-time event stream:
- A Live Feed of every filing as it hits EDGAR (no signup required)
- Sentinels that match new filings against your tickers, themes, and rules
- AI Briefings that summarize what just happened and why it matters
- Webhooks and an MCP server so your tools and your agents stay in sync
If you've ever written a script that polls EDGAR every minute, you've already half-built edgar.tools. The platform finishes the job.
I'm Dwight Gunning. I started edgartools because corporate reporting is fascinating and the tools for working with it weren't. The library, the blog, and the platform are the answer that grew out of scratching that itch.