What problem does it solve?
Fleet Control removes the need for manual oversight by continuously monitoring managed Aeon instances, detecting degradation, and preparing concrete next actions based on repository and workflow signals.
Core Features & Use Cases
- Fleet-wide health monitoring: Health checks GitHub repo metadata, 24h workflow runs, and child cron-state, then classifies instances into healthy, warning, degraded, stale, pending, unreachable, or archived.
- Decision-ready reporting: Produces a verdict-first output (plus deltas vs prior state) suitable for operators who need scanable, action-oriented summaries.
- Automated dispatching: Dispatches a specified skill to targeted instances (healthy/warning/degraded) after verifying the child repo contains the requested skill and optionally checking enabled status in aeon.yml.
- Operational guardrails: Enforces pre-flight checks for GitHub auth and API rate limits, bootstraps missing registries/state safely, and logs outcomes to repository memory paths.
Quick Start
Run Fleet Control with an empty command to perform the default health check and generate fleet-level actions and delta reporting for today.