llms.txt & docs for agents

Notiflows publishes its documentation in machine-readable forms so you can drop authoritative context into an AI assistant instead of relying on its training data. Everything here is generated from the same docs you're reading, so it never drifts.

/llms.txt

A concise, link-first index of the documentation, following the llms.txt convention. It lists every doc page (grouped by section) with a one-line description and a link, so an assistant — or you — can quickly find the right page and fetch it on demand.

https://notiflows.com/llms.txt

Use it as the lightweight entry point: hand it to an assistant that can follow links, and it will pull the specific pages it needs rather than ingesting everything.

/llms-full.txt

The entire documentation concatenated into one plain-text file — every page's title, description, source URL, and full markdown body. Use it when you want to load the complete docs into a model's context in a single fetch (for a coding agent, a RAG index, or a one-shot prompt).

https://notiflows.com/llms-full.txt

The generated API reference is intentionally excluded from llms-full.txt because those pages are component tags rather than readable prose — point agents at the API reference directly for endpoint details.

Raw markdown per page

Every doc page is also available as raw markdown — just append .md to any docs URL:

https://notiflows.com/docs/cli/commands.md
https://notiflows.com/docs/concepts/notiflows.md

This serves the page's underlying markdown as text/markdown, so you can fetch a single page directly into an assistant's context (the "copy for LLM" and "view markdown" actions at the top of each page use the same source). Use it when you want one page rather than the whole llms-full.txt.

Related

• Notiflows for AI — the AI surfaces overview.
• Agent skills — installable, grounded context packets for coding agents.
• MCP server — connect an AI editor directly to the Management API.
• Agent toolkit — turn notiflows into typed tools for your own agents.
• API Reference — the interactive REST reference (fetch directly, not via llms-full.txt).