LLM-Ready Docs#
Every page on this docs site is also available as structured, JSX-stripped markdown for AI tools.
llms.txt - canonical index#
Following the llmstxt.org standard: a plain-text index with one line per page as - [Title](https://tetrafi.io/docs/<slug>): <description>, grouped by section.
- Endpoint:
https://tetrafi.io/llms.txt - Content-Type:
text/plain - Cached for 1 hour with
stale-while-revalidatefor a day.
Consumers: Claude, ChatGPT, Cursor, Phind, and any other tool that reads /llms.txt from a site root.
Raw markdown for any doc#
Every doc is reachable at /docs/raw/<slug> as JSX-stripped markdown.
- Endpoint pattern:
https://tetrafi.io/docs/raw/<section>/<page> - Content-Type:
text/markdown - Components are transformed to their closest markdown equivalent:
<CodeBlock>→ fenced block,<Callout>→ blockquote,<Term>→ plain text,<Badge>→[label].
Examples#
/docs/raw/overview/concepts- core concepts page as markdown/docs/raw/architecture/quote-pipeline- quote pipeline as markdown/docs/raw/integration/api-reference- REST API reference as markdown
Why#
AI tools, LLM-aware IDEs, and code assistants increasingly expect a machine-readable docs surface. HTML + MDX bundles are lossy for them; structured markdown is not.
The raw transform is best-effort. If you need perfect fidelity for a specific component, the authoritative source is the rendered HTML page - use it alongside the markdown feed for disambiguation.