How We're Preparing Our Products for the Future of AI

We've rolled out llms.txt files across our websites to help large language models better understand our products. That means more accurate answers, smarter agents, and a better experience for developers working with Iron Software libraries.

As AI tools like ChatGPT, Google Gemini, and Claude reshape how developers find and use technical information, we at Iron Software are making key improvements to ensure our product documentation and developer experience are LLM-ready.

We've introduced new llms.txt files across our product sites - a Markdown-based format designed to help AI agents understand, navigate, and reference our tools more accurately and efficiently.

What Is llms.txt?

llms.txt is a new open format proposed by Jeremy Howard (co-founder of Fast.ai) that provides a clean, structured overview of the most important pages on a website - written specifically for large language models (LLMs) and AI agents.

While traditional HTML pages often include layout noise, complex JavaScript, and navigation-heavy structures, llms.txt allows developers and AI tools to quickly locate:

• API documentation
• Code examples and tutorials
• Licensing and EULA details
• Feature lists and changelogs
• Troubleshooting pages

The format is Markdown-based, human-readable, and machine-friendly - the ideal mix for both developers and inference-time LLMs.

Where to Find Our llms.txt Files

We've created dedicated llms.txt files for every major Iron Software product, including both .NET and Python libraries.

Here's where you can find them:

Why This Helps Developers (and AI)

These new llms.txt files ensure that:

• ChatGPT and other LLMs return more accurate answers about our tools
• Key documentation isn't buried in navigation-heavy HTML
• Code examples and licensing terms are properly represented in AI-generated content
• AI agents (e.g., in dev tools or IDEs) can index our offerings efficiently

For example, instead of parsing a full HTML page for a licensing clause or code snippet, an AI agent can now just load a clean Markdown file and extract the relevant entry - no scraping, guessing, or hallucinating.

What's Next?

We're not stopping here.

In the coming months, we plan to:

• Support AI-oriented chunking for future retrieval-augmented generation (RAG) tasks
• Open up feedback for developers using AI to work with our tools

If you're using AI tools in your workflow and want to collaborate or share feedback - reach out to us.

Resources

• Learn about the llms.txt format
• Contact Support