On-call rotation
Three founders rotate weekly. Pager goes through Discord @oncall mention (Year-1 free tier; PagerDuty free tier at 5 users is the Year-2 upgrade path).
Rotation schedule
| Week starting | Primary | Secondary | Tertiary |
|---|---|---|---|
| 2026-W22 (May 25) | F1 | F2 | F3 |
| 2026-W23 (Jun 01) | F2 | F3 | F1 |
| 2026-W24 (Jun 08) | F3 | F1 | F2 |
| 2026-W25 (Jun 15) | F1 | F2 | F3 |
Rotate weekly on Monday at 00:00 UTC. Buildathon week (2026-W23) F3 covers demo rehearsals; F1 covers infra; F2 covers comms.
Page sources
- Better Stack uptime any monitor red for 60 seconds
- Sentry any unresolved issue spike (more than 10 events in 5 min)
- Discord
#alertschannel user-reported issue with screenshot - GitHub Actions failure on main CI workflow red on push
Response SLAs
| Severity | Acknowledge | First update | Resolved |
|---|---|---|---|
| SEV-0 (funds at risk OR false claim shipped) | 5 min | 15 min | 1 h |
| SEV-1 (core flow down) | 15 min | 30 min | 4 h |
| SEV-2 (degraded) | 1 h | 4 h | 24 h |
| SEV-3 (cosmetic) | next business day | next business day | next sprint |
During an incident
- Acknowledge in Discord
#alertswith the incident ID + SEV. - Pull the relevant runbook from
runbooks/incident-*.md. - Status update every 30 min until resolution.
- After resolution: file
/incidents/<date>-<slug>.mdwithin 7 days. - Add regression test or alert tuning before closing.
Escalation
- If primary is unreachable for 15 min, secondary takes over
- If secondary is unreachable for 15 min, tertiary takes over
- All three unreachable: SEV-0 pause every contract via Praetor
emergency-pause (instant, multisig-only). Resume after the rotation recovers.
Founder contacts
- F1: f1@useatrium.me . Discord: @F1 . Phone: in 1Password
- F2: TBD per cohort founder lineup
- F3: TBD per cohort founder lineup
Phone numbers + emergency contacts live in the founder 1Password vault, NOT in this repo.