llms.txt#

OnchainOS Market llms.txt Structure#

llms.txt is a proposed standard designed to help AI language models efficiently understand and navigate documentation or websites. It enables Agents to quickly grasp the structure of documentation and locate relevant pages. Based on minimal context, it can precisely identify the appropriate module within shorter response times, reduce hallucinations, and lower token consumption—serving as a low-cost navigation layer.

A typical llms.txt file includes:

• Site title (H1)
• Sections organized by module (H2)
• A link to each page + a one-sentence description (used for routing and retrieval)

In addition to llms.txt, we also provide llms-full.txt. Unlike a directory-style index, it aggregates the full site documentation in Markdown format (including more granular descriptions and examples) for deeper indexing and search.

OnchainOS Market llms.full.txt Structure#

Suitable for:

• AI tools that require full-context access (deep indexing / advanced search / offline knowledge base construction)
• Developers who want a comprehensive view of all pages and resources at once
• Building custom AI workflows (e.g., full indexing first, then chunk-based retrieval as needed)