LLM and MCP Readiness
Voicetta documentation is designed to be readable by humans and by LLMs.
Why it matters for your business
- Save money: accurate crawler-readable docs reduce repeated support and sales explanations.
- Work faster: humans and robots can retrieve the same setup guidance.
- Standardize service: public docs become the shared source for app behavior and Guest language.
- Delight every Guest: teams and AI tools learn the same Guest-first operating standard.
The goal is simple:
- Humans should get a clean documentation experience.
- LLMs should get stable Markdown.
- Crawlers should discover docs through sitemap and
llms.txt. - MCP documentation should be added only after tool contracts are stable.
Markdown Access
Every public documentation page should have a Markdown version:
- HTML:
/docs/:slug - Markdown:
/docs/:slug.md
The full documentation bundle is available as Markdown:
/docs.md
The LLM discovery file is:
/llms.txt
Sitemap
Sitemap is the primary discovery mechanism.
The sitemap should include:
/docs/docs/:slug/docs.mdwhen full Markdown should be discoverable
Individual .md pages may also be included if we want search engines and AI crawlers to discover them directly.
llms.txt
llms.txt should list the most useful Markdown resources:
- Full documentation Markdown
- Documentation index
- Changelog
- Release notes
- Agent tools
- MCP readiness
MCP Roadmap
MCP should be documented in stages:
- Document available agent tools.
- Document future MCP concepts and auth boundaries.
- Publish stable tool descriptors only after the product contract is clear.
- Keep docs, tool names, and examples synchronized.
Rule
Do not document a tool as available unless Voicetta can actually perform the action.