> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mov-eat.app/llms.txt
> Use this file to discover all available pages before exploring further.

# Runbooks

> Short operational procedures for common Moveat incidents and maintenance tasks.

# Runbooks

This page lists common operational actions for Moveat.

## Platform does not start

Check:

1. Container logs.
2. Missing environment variables.
3. Database availability.
4. Redis availability.
5. Prisma migrations.
6. Nest dependency injection errors.

If the error is a missing provider in a module, the likely fix is importing the module that exports that provider.

## Agent does not pick up new environment variables

Restart/recreate the Agent container after changing the production environment file.

A simple container restart may not be enough if the compose environment was not recreated.

## Database password changed

If the Postgres role password changes, update the production environment file and recreate affected services.

Verify with `psql` through the local bound port or via Docker network.

## Grafana dashboard looks empty

Check:

1. Time range.
2. Loki datasource health.
3. Alloy logs.
4. Container labels.
5. Whether the selected panel filters exclude the service.

## Logs are too noisy

Health check logs should be filtered out in dashboards or at the log collection layer.

If disk usage grows quickly, check Docker log rotation and Loki retention.

## Agent cannot call Platform

Check:

1. Platform container is healthy.
2. Agent has Platform base URL configured.
3. Internal service token matches on both sides.
4. Caddy routing is not involved for internal Docker-network calls unless intentionally configured.
5. Platform internal endpoint logs show the request.

## Documentation changes do not appear

Check:

1. Changes were pushed to the docs repo.
2. Mintlify GitHub integration is connected.
3. Mintlify deployment succeeded.
4. DNS points to the correct Mintlify target.
5. Browser cache is not showing an old page.