# Fix: /llms-full.txt present (single-fetch full corpus)

> Web-audit fix skill for the `llms-full-txt` check (Content for agents, MAY).

## Goal

Serve the whole docs corpus as markdown at /llms-full.txt for one-fetch ingestion.

## Fix

For a docs site, concatenate every page as markdown at `/llms-full.txt` so an agent can ingest
the whole corpus in one fetch instead of crawling. Keep it generated from source so it never
drifts from the live pages.

## Resources

- [llmstxt.org](https://llmstxt.org/)

## Copy-paste prompt

Paste this into your coding agent, replacing the Issue line with the Result line from [your audit](https://anc.dev/web-audit):

```text
Goal: Serve the whole docs corpus as markdown at /llms-full.txt for one-fetch ingestion
Issue: <the audit's finding for this check>
Fix: For a docs site, concatenate every page as markdown at `/llms-full.txt` so an agent can ingest the whole corpus in one fetch instead of crawling. Keep it generated from source so it never drifts from the live pages.
Skill: https://anc.dev/web-audit/skill/llms-full-txt
Docs: https://llmstxt.org/
```

## Verify

Re-run the audit at [https://anc.dev/web-audit](https://anc.dev/web-audit) or call the `audit_website` MCP tool; the `llms-full-txt` check should report `pass`.
