Skip to content
Community226

LLMs.txt

This is the one AI-access mechanism on this site that’s real and live today — not planned, not conceptual. llms.txt is a convention that lets AI tools discover a site’s content in a compact, machine-readable form instead of parsing rendered HTML.

An llms.txt file is a plain-text index of a site’s documentation, generated for AI consumption rather than human browsing. This site generates its own automatically at build time — real content, not a hand-maintained list that can drift out of sync with what’s actually published.

Three files, each a different level of detail:

  • /llms.txt — a compact index of the documentation set, linking out to the fuller versions.
  • /llms-full.txt — the complete documentation content in one machine-readable file.
  • /llms-small.txt — an abridged version with non-essential content removed, for tools with tighter context budgets.

All three are generated from this site’s real, published content — the same pages you’re browsing right now, not a separate maintained copy. Local-only reference/demo pages are stripped out of all three, so an agent reading them only ever sees real SOLMEME documentation.

Beyond the three index files, any individual documentation page on this site can be requested as plain Markdown by appending .md to its URL — useful for an agent that wants one specific page’s content rather than the entire index.

Point any AI tool or agent that supports the llms.txt convention at /llms.txt as this site’s entry point. For an agent consuming this site’s content, Knowledge Policy governs how to treat what it finds there — what counts as canonical, how to judge freshness, and when to cite a source rather than state something as settled fact.