ralph-loop-control-plane v1.0
What it does
ralph-loop-control-plane reviews JSON plans for long-running autonomous loops before they go live. It validates cadence, retry settings, stop conditions, escalation rules, and human review checkpoints, then produces a clean summary operators can act on.
Use cases
- Automation builders running recurring loops: validate that a planned workflow has proper stop conditions and does not retry forever.
- Operations teams approving autonomous jobs: get a short, readable summary of loop safeguards before sign-off.
- Anyone hardening agent autonomy: spot missing escalation paths and weak control settings before a loop causes trouble.
Requirements
- Dependencies: Python 3 only, standard library modules only
- API keys: None
- OpenClaw version: Compatible with OpenClaw v2026.3.23+
Example usage
python3 scripts/ralph_loop_control_plane.py --plan /tmp/loop-plan.json --check
Expected output
Plan validation passed.
With incomplete plans, it prints severity-ranked findings such as missing required keys, unsafe retry behaviour, or absent human review thresholds.
Price
$4.99