mirror of
https://github.com/hanzoskill/otter.git
synced 2026-10-30 21:13:01 +00:00
2.0 KiB
2.0 KiB
Twenty CRM Integration
Sync Otter.ai transcripts to Twenty CRM as notes on contacts or companies.
Setup
Set environment variables:
export TWENTY_API_URL="https://api.your-twenty-instance.com"
export TWENTY_API_TOKEN="your-api-token"
Workflow
1. Get Transcript Summary
uv run skills/otter/scripts/otter.py summary <speech_id>
2. Sync to Contact
uv run skills/otter/scripts/otter.py sync-twenty <speech_id> --contact "John Smith"
3. Sync to Company
uv run skills/otter/scripts/otter.py sync-twenty <speech_id> --company "Acme Corp"
Agent Integration
When the sync-twenty command runs, it outputs JSON that the agent should process:
{
"action": "sync_to_twenty",
"title": "Weekly Standup",
"speech_id": "abc123",
"transcript_text": "...",
"contact": "John Smith",
"company": null
}
The agent should then:
- Summarize the transcript (key points, action items)
- Find the contact/company in Twenty CRM
- Create a note with the summary
- Optionally attach full transcript
Twenty API Endpoints Used
POST /api/notes- Create noteGET /api/people?filter[name]- Find contactGET /api/companies?filter[name]- Find company
Example Note Format
## Meeting Summary: Weekly Standup
**Date:** 2026-01-06
**Duration:** 45 minutes
**Participants:** John, Sarah, Mike
### Key Points
- Project timeline on track
- Budget review needed
- New feature request from client
### Action Items
- [ ] John: Send budget proposal by Friday
- [ ] Sarah: Schedule client demo
### Full Transcript
[Attached as separate note or linked to Otter]
Automatic Sync (Cron)
Set up a cron job to check for new transcripts and auto-sync:
{
"name": "otter-sync",
"schedule": {"kind": "cron", "expr": "*/30 * * * *"},
"payload": {
"kind": "agentTurn",
"message": "Check Otter for new transcripts. If found, summarize and ask if I should sync to Twenty CRM."
}
}