Honeybadger Setup

Phase 12 observability, heartbeat monitoring for all Atrium daemons.

Prerequisites

• Honeybadger account (Small plan via GitHub Student Pack, see runbooks/student-pack-setup.md)
• Doppler access

Steps

1. Activate Honeybadger

1. Sign up at honeybadger.io or activate via Student Pack.
2. Create project atrium-daemons.

2. Create Heartbeat Check-Ins

Create a check-in for each daemon:

For each, note the check-in URL (format: https://api.honeybadger.io/v1/check_in/<TOKEN>).

3. Add to Doppler

The heartbeat helper uses a single env var with a <NAME> placeholder:

# Template URL, each daemon replaces <NAME> at runtime
doppler secrets set HONEYBADGER_HEARTBEAT_URL="https://api.honeybadger.io/v1/check_in/<TOKEN_FOR_NAME>"

Alternatively, use per-daemon URLs:

doppler secrets set HONEYBADGER_HEARTBEAT_URL_NOTIFIER="https://api.honeybadger.io/v1/check_in/<notifier-token>"
doppler secrets set HONEYBADGER_HEARTBEAT_URL_VIGIL="https://api.honeybadger.io/v1/check_in/<vigil-token>"
doppler secrets set HONEYBADGER_HEARTBEAT_URL_LANTERN="https://api.honeybadger.io/v1/check_in/<lantern-token>"

If using per-daemon URLs, update heartbeat.ts in each service to read the specific var.

4. Configure Miss Alerts

In Honeybadger UI → Check-Ins → each check-in:

• Alert if no ping received in 2× expected interval
• Notification: Discord webhook (add in Honeybadger → Integrations → Webhook)

5. Verify

1. Start each daemon locally with the env var set.
2. Confirm pings appear in Honeybadger → Check-Ins within one interval.
3. Stop a daemon; confirm alert fires after grace period.