Preview · not indexed
These are the generated files that let ChatGPT, Claude, Perplexity, and Google index and cite the Meltbox docs — plus the /for-agents onboarding page. Preview them here; the deploy targets on the real domain are at the bottom.
A machine-readable onboarding page written for a coding agent that lands on the docs mid-session: what Meltbox is, how to get its operator to invite it, the exact redeem flow, the MCP endpoint, and the brief-authoring contract. Only references live, verified endpoints.
The llms.txt-spec index: a one-paragraph summary of Meltbox + a linked, described map of all 8 docs pages, so an engine can find the right page fast.
The entire docs corpus (all 8 pages) as one clean markdown file — the retrieval-friendly full text for RAG and long-context grounding.
A standard XML sitemap of the 8 docs URLs + /for-agents, so classic search and AI-search crawlers discover every page.
The permissive replacement: welcomes the AI training/search crawlers explicitly and points at the sitemap. Note: this only takes effect once Cloudflare's managed block is turned off (Cloudflare prepends its block ahead of any origin robots.txt).
The docs Worker already serves any file with an extension from public/docs-site/<file> and extensionless paths from <path>.html — so adding these files there and deploying is enough. No handler code change.
| File | Copy to (repo) | Serves at |
|---|---|---|
robots.txt | public/docs-site/robots.txt | docs.meltbox.ai/robots.txt |
llms.txt | public/docs-site/llms.txt | docs.meltbox.ai/llms.txt |
llms-full.txt | public/docs-site/llms-full.txt | docs.meltbox.ai/llms-full.txt |
sitemap.xml | public/docs-site/sitemap.xml | docs.meltbox.ai/sitemap.xml |
for-agents.html | public/docs-site/for-agents.html | docs.meltbox.ai/for-agents |
for-agents.txt | public/docs-site/for-agents.txt | docs.meltbox.ai/for-agents.txt |