# Ingest LLM-friendly documentation from any site that supports the llms.txt standard

> LLMsTxtTools and LLMsTxtReader add native support for the llms.txt standard — a Markdown-based file that websites publish at /llms.txt to provide LLMs with a concise, structured index of their…

- Published: 2026-04-10
- Author: Agno Team
- Category: Changelog
- Canonical: https://agno-com-nine.vercel.app/articles/ingest-llm-friendly-documentation-from-any-site-that-supports-the-llms-txt-standard
- Markdown: https://agno-com-nine.vercel.app/articles/ingest-llm-friendly-documentation-from-any-site-that-supports-the-llms-txt-standard.md

`LLMsTxtTools` and `LLMsTxtReader` add native support for the `llms.txt` standard — a Markdown-based file that websites publish at `/llms.txt` to provide LLMs with a concise, structured index of their documentation, free of navigation elements, JavaScript, and other noise that wastes context. Agents can now fetch, read, and work with `llms.txt` files directly, making it straightforward to build agents that are grounded in up-to-date third-party documentation without manual content pipelines.

Details:

- `LLMsTxtReader` ingests any `llms.txt` file into a knowledge base for retrieval and RAG
- `LLMsTxtTools` lets agents fetch and query `llms.txt` indexes directly as a tool call
- Compatible with any site publishing the standard, including `https://docs.agno.com/llms.txt`
- No preprocessing required — `llms.txt` files are already structured for LLM consumption

See [cookbook](https://github.com/agno-agi/agno/blob/main/cookbook/91_tools/llms_txt_tools.py) for reference