Prism change request: #1304
Original request
Implement the approved AI Agent Readiness Level 3 package for https://www.raidguild.org. Allow search and AI-answer use, prohibit training via Content Signals, expose a truthful llms.txt, advertise agent resources through Link headers, and provide same-URL Markdown content negotiation for public pages. Add focused automated coverage and run repo-native validation. Open a PR only; do not merge or deploy production without human review.
Triage summary
The Next.js 15 App Router site already has public/llms.txt (added in PR #96), canonical metadata, robots and sitemap metadata routes, and public pages at / and /join plus the legacy /witch route. Remaining work centers on an explicit crawler/content-signal robots policy, response Link headers that preserve existing Link values, Accept: text/markdown negotiation for public content only, and an automated test harness/fixtures for these behaviors. Target branch is main; implementation branch is codex/cr-1304-raidguild-org-ai-agent-readiness-level-3.
Initial fix notes
- Replace or extend the generated robots response in a way that can emit exact User-agent and Content-Signal directives while retaining wildcard allow, host, and sitemap declarations.
- Add reusable Markdown representations for public routes and middleware/route handling that excludes API, assets, framework internals, and non-public surfaces; set Content-Type and Vary: Accept correctly.
- Add llms.txt and sitemap Link relations to HTML responses without overwriting framework preload Link headers.
- Validate normal HTML behavior, exact Markdown Accept behavior, robots directives, llms.txt truthfulness, and header preservation.
- Run lint, TypeScript checking, tests, and production build. Create a PR for human review; live-domain and isitagentready.com verification happen only after deployment.
Prism change request: #1304
Original request
Implement the approved AI Agent Readiness Level 3 package for https://www.raidguild.org. Allow search and AI-answer use, prohibit training via Content Signals, expose a truthful llms.txt, advertise agent resources through Link headers, and provide same-URL Markdown content negotiation for public pages. Add focused automated coverage and run repo-native validation. Open a PR only; do not merge or deploy production without human review.
Triage summary
The Next.js 15 App Router site already has public/llms.txt (added in PR #96), canonical metadata, robots and sitemap metadata routes, and public pages at / and /join plus the legacy /witch route. Remaining work centers on an explicit crawler/content-signal robots policy, response Link headers that preserve existing Link values, Accept: text/markdown negotiation for public content only, and an automated test harness/fixtures for these behaviors. Target branch is main; implementation branch is codex/cr-1304-raidguild-org-ai-agent-readiness-level-3.
Initial fix notes