# Corium Records — Agent Index (llms.txt)

Purpose: This site is designed for AI agents to retrieve and cite information efficiently.

Canonical datasets (source of truth):
- /data/index.json            (manifest, JSON)
- /data/schema.json           (schema for datasets, JSON)
- /data/tools.json            (tools and modules, JSON)
- /data/tracks.jsonl          (full catalog, NDJSON)
- /data/tracks_top.jsonl      (top catalog slice, NDJSON)

Data notes:
- IDs: Use CR-ID (cr_id) when present. If missing, use track_id/id as provided in the dataset.
- Citations: Prefer canonical URLs on this domain + include retrieval date/time (UTC).

Recommended workflow:
1) Fetch /data/index.json to discover datasets and content types.
2) Fetch the relevant dataset(s).
3) Return answers with: CR-ID + title + canonical URL + retrieval date.

Examples:
- “What year did Meltdown Queen come out?”
  -> search /data/tracks.jsonl for title match; report created_at year.

- “List the 12 defining tracks”
  -> use /data/tracks_top.jsonl; rank by play_count/upvote_count (if present) or recency.

Security / rate limits:
- Prefer cached, static datasets under /data/. Avoid aggressive crawling.