OUTCOMES
What this Prompt helps you deliver
- Critical services are observable
- Backups restore successfully
- Upgrades have a rollback path
BEFORE YOU START
Before you start
- A deployment already runs
- Retention and recovery objectives are defined
What AI should inspect first
deploymentdocker-compose.deploy.ymlapps/tabtin_djangodocs/agent/evidence-map.mdREADY TO RUN
Complete Prompt
Copy the complete Prompt with its execution and acceptance boundaries.
Turn a running TabTin deployment into a production system that can detect failure, find causes, restore data, and upgrade safely—not a pile of monitoring products. Read deployment, docker-compose.deploy.yml, apps/tabtin_django, and docs/agent/evidence-map.md. Establish topology, key user flows, data locations, downtime, recovery point, and alert recipients. Reuse current logs, metrics, backup, and upgrade scripts before adding complexity.
Define health and failure signals for Web, API, database, cache, collaboration, realtime, and background jobs. Standardize structured logs, request correlation, and redaction while retaining user-flow evidence. Schedule encrypted off-site backups for database, uploads, and critical configuration with retention. Define upgrade prechecks, migrations, service order, health verification, and rollback. Alert only on actionable capacity, error rate, queue backlog, disk, and certificate expiry.
Cause one safe non-production fault and prove alerts/logs identify the service. Restore a backup in isolation. Rehearse an upgrade and critical user flows. Confirm no password, token, or sensitive body appears in logs, backups, or alerts. Report signals, thresholds, log locations, restore evidence, upgrade/rollback, daily duties, and which checks still need a production window.