CLI - Apgar

The apgar command is the entry point of the API and worker image. In the compose stack, run it inside a running container:

docker compose exec apgar-api apgar --version

With the Helm chart, run it in the API pod:

kubectl -n apgar exec deploy/apgar-api -- apgar --version

Commands

Command	Effect
`apgar migrate`	Create the database if it is missing and apply pending migrations. A second run changes nothing
`apgar api`	Serve the HTTP API on `APGAR_API_HOST:APGAR_API_PORT`
`apgar worker`	Run the worker. Run exactly one
`apgar demo seed`	Insert the scripted demo agents and their verdicts
`apgar demo purge`	Delete the demo agents and every row that belongs to them
`apgar --version`	Print the version

apgar api and apgar worker check the schema version before they start.

demo seed

Option	Default	Meaning
`--days`	`14`	Days of scripted traffic
`--turns-per-day`	`300`	Turns per day
`--seed`	`42`	Random seed
`--purge-first`	off	Delete existing demo data first

When demo data already exists, the command warns and seeds on top of it; pass --purge-first to replace it. Scheduled checks skip demo agents; their verdicts are computed once, while seeding. The Seed demo data button on the empty Agents page queues the same seeding in the worker.

Exit codes

Code	Meaning
`0`	Success
`1`	Startup failed, for example ClickHouse is unreachable or another worker is running
`2`	Invalid configuration: an environment variable is missing or invalid
`3`	The database schema version does not match this version of Apgar. Run `apgar migrate`. `api` and `worker` report it this way; the demo commands fail with a traceback and code 1