| name | meeting |
| description | Create a meeting transcript file in library/meetings/. Collects title, date, project, participants — then creates the structured file so you can paste the transcript directly into it. |
Meeting
Create a structured meeting file in library/meetings/[project]/ ready for transcript paste.
Workflow
Step 1: Collect metadata
Ask (all in one message, 3 questions):
- Meeting — title (e.g. "Alessia weekly", "Founder check-in")
- Project — which project/context? (type a project name)
- Participants — who was in the meeting? (comma-separated names)
Date defaults to today unless user specifies otherwise (e.g. "yesterday's meeting with Alessia").
Step 2: Resolve GCal event ID
Scan local events/*.md files for events on the meeting date:
Glob events/*.md
For each file, read the frontmatter and check if start_time matches the meeting date (YYYY-MM-DD prefix).
- One match → use its
gcal_event_id silently
- Multiple matches → show titles, ask which one (or "none")
- No matches → skip, leave
gcal_event_id out of frontmatter
Step 3: Build file path
library/meetings/[project]/YYYY-MM-DD-[slug].md
Slug: lowercase title, spaces→hyphens, strip special chars. Max 40 chars.
Examples:
library/meetings/[project-a]/2026-02-26-weekly-sync.md
library/meetings/[project-b]/2026-02-26-founder-check-in.md
Create directory if it doesn't exist (Write tool handles this automatically).
Step 4: Write file
Use Write tool to create the file:
---
type: meeting
date: YYYY-MM-DD
project: [project]
participants: [name1, name2]
gcal_event_id: [id if resolved]
duration_min:
---
Leave body empty — user will paste transcript directly.
duration_min is always included as empty field — user can fill it or leave blank.
Only include gcal_event_id line if resolved.
Step 5: Confirm
Meeting file ready:
library/meetings/[project]/YYYY-MM-DD-[slug].md
Open the file and paste your transcript below the frontmatter.
The intelligence pipeline will tag and embed it automatically.