David decided: Added anti-fabrication constraint to MCP connector audit

David kicked off an 8-agent audit of the granular MCP connector against an ideal end state, but embedded a hard constraint from the start:

run 8 agents (opus 4.7 max) to go and audit the app and mcp connector against the ideal end state where the mcp connector lets the user's AI fully manipulate the full state of their profile, every primitive, every section etc and come back with a final plan. the constraint must be that the system design cannot allow the user's AI to lie on their behalf.

He also added a product question alongside the audit: whether the system could fingerprint which AI host (Claude Code, Claude Chat, Codex, ChatGPT, etc.) is writing to the MCP server, to enable deterministic AI-tool-usage tracking on the profile. The anti-fabrication framing shaped the entire design direction that followed.