Add billing to keep writing trusted context after the trial.
How to use the catalog
This is an AI-native data catalog for Claude Code, Cursor, and Codex: metric definitions, source routes, answer paths, explanatory frames, proof receipts, caveats, decisions, playbooks, and evals. The strongest source routes point to a customer-owned GitHub SQL/dbt source or a dashboard screenshot/URL with a number tie-out. Claude Code, Cursor, and Codex read it through MCP; the portal shows the current approved entries and review state.
In Claude Code, use /chatdata:start for the full feature map, /chatdata:catalog for catalog guidance, /chatdata:sync-context to save reusable context, and /chatdata:publish-patch to approve reviewed updates.
Approved catalog entries
Loading approved catalog entries from ChatData Context Hub.
Loading approved context files.
Catalog artifact detail
Select an approved catalog entry to inspect it.
definitions
Trusted metrics
Approved business definitions that AI tools can cite without guessing.
Metadata-only by design, with guards for obvious secrets and raw-row patterns. Stored in ChatData Context Hub, US region. Revoke anytime from Settings.
Consent status has not changed in this dialog yet.
ChatData stores analytics metadata in the Context Hub: metric definitions, SQL patterns, answer paths, caveats, review states, and proof notes. The content guard blocks obvious secrets and raw-row patterns before writes are accepted.