DevOps

Use flowbot devops for capability devops.
CLI root is devops — do not invent flowbot devops unless cli.md lists it as an alias.
Prefer the workflows below; load references/cli.md only when you need a flag or subcommand not covered here.

CLI limits: There is no flowbot devops health; use flowbot devops status for aggregate readiness, then backend-specific health/metrics commands.

JSON fields: Prefer -o json. Start with devops status to see which backends are configured.

Setup

1. Ensure CLI auth: flowbot login
2. Set server via FLOWBOT_SERVER_URL or --server-url; optional --profile, --debug / -d
3. Prefer -o json when parsing results programmatically
4. Destructive commands often need -y / --yes in non-interactive sessions — check cli.md
5. Token scopes: service:devops:read

Workflows

Check which backends are configured

When a user asks what devops tools are available:

1. flowbot devops status
2. Only call subcommands for backends reported as configured.

Inspect monitoring and routing

When a user wants a quick infrastructure snapshot:

1. flowbot devops beszel systems
2. flowbot devops uptimekuma health
3. flowbot devops traefik routers
4. flowbot devops grafana health
5. flowbot devops wakapi projects
6. flowbot devops dozzle health
7. flowbot devops netalertx totals

Query observability backends via Grafana

When a user wants metrics, logs, traces, or profiles:

1. flowbot devops grafana datasources
2. flowbot devops grafana query --backend prometheus --expr 'up'
3. Use backend alloy|loki|tempo|pyroscope with the matching expression language; prefer -o json for parsing.

Troubleshooting