A clean operating brief from old chats
Ask your Instinct to build a short operating brief from your earlier assistant chats: role, goals, recurring duties, communication style, what to bring back for your approval, and what you do not want. Facts and guesses are kept apart, and you correct the draft before anything is saved.
The prompt
Help me write a short operating brief from my prior assistant conversations: my role, current goals, recurring responsibilities, communication style, decisions to bring back for review, and things I do not want. Separate confirmed facts from guesses, let me correct it, and strip private details I do not want imported. Share only the cleaned brief; do not turn guesses into standing assumptions.
Prerequisites
- Access to the earlier conversations: an export or pasted chats from a previous assistant.
- A few minutes to read the draft and correct it.
Permission limits
- Reads only the chats you hand it.
- Strips private details you asked not to import and shares only the cleaned brief.
- Guesses never become standing assumptions. Only what you confirm is kept.
Test status
Not tested by Builder. The tip comes from a LinkedIn post by Eynav, not from a run of ours.
Result: No run result. Status as of 2026-10-09.
Known failure: No failure recorded in a run. The known risk the prompt is built around: a guess presented as fact and saved as a standing rule. That is why it has a separate-and-correct step.
Source: Eynav on LinkedIn · Updated