Monitoring that validates protocol functionality and alerts when things break.
| Test suite | Status | Resources | Run frequency | Networks |
|---|---|---|---|---|
| Performance |  |
Workflow / Test code | Every 30 min | dev production |
| Delivery |  |
Workflow / Test code | Every 30 min | dev production |
| Browser |  |
Workflow / Test code | Every 30 min | dev production |
Location: monitoring/performance.test.ts
- Timing/throughput for core ops (create, canMessage, inboxState, DM send/stream)
- Scaled group ops per size (create, sync, add/remove members, metadata updates)
- Scalability checks across 50–250 members (create, sync, stream, metadata)
Location: monitoring/delivery.test.ts
- Message delivery and order via streams and polling
- Recovery after stream interruptions
- Tests with regressions from the last 3 versions
Location: monitoring/browser/browser.test.ts
- Core conversation + message streaming flows in browser SDK (Playwright)
Location: monitoring/networkchaos
- Resilience under partitions/delays/failures on a local 4-node cluster
- Includes smoke tests, DM duplicate prevention, group reconciliation, client partition, key rotation
- Performance Tests: 1,440 writes/day (group creates, sends, updates, adds/removes across sizes)
- Delivery Tests: 7,224 writes/day (group creates + 150 sends per run across 3 tests)
- Agents-DMs Tests: 5,400 writes/day (DM creates + sends with retries per agent)
- Total Daily Writes: ~14,064