Meetings

The meeting ends. The memory updates.

A meeting record is a `meeting` object: a title, the date it was held, the company it concerns, and the notes or transcript as its body. That alone gives you a searchable meeting log your AI can read.

The click that matters is Extract: the model reads the notes and proposes what memory should learn — decisions, dates, people, follow-ups. Everything lands in your review queue first; approve what's right, and the meeting has fed your CRM and your task board without retyping a line.

What it does

Log meetings with date, company and full notes or transcript.
One-click extraction turns notes into proposed facts — reviewed by you.
Follow-ups extraction finds become todos on the Tasks board.
Ask your AI what was decided in any meeting — same records.

What it stores

As ordinary objects in your workspace memory — readable by your AI over MCP, browsable like any memory.

meeting — A meeting that took place — held_on is its date; the notes or transcript live in the description.
company — A business you deal with — customer, prospect or partner.