What problem does it solve? Adding scheduled jobs to OpenClaw can break the symlinked cron state or leak context between runs. This Skill provides a validated, repeatable procedure for registering cron, every, or at jobs in openclaw.json while protecting the transient jobs.json state and gateway deployment. ## Core Features & Use Cases - Validated Job Registration: Enforces required inputs (job_id, agent_id, schedule_type, schedule_value, prompt) and rejects invalid schedule types before writing config. - Isolated Execution by Default: Jobs run in isolated sessions to prevent context bleed and control token cost across scheduled tasks. - Safe Deployment Workflow: Removes the transient ~/.openclaw/cron/jobs.json, deploys with stow, restarts the gateway via launchctl, and verifies job materialization and scheduler logs. - Use Case: Schedule a recurring nightly memory-maintenance agent job by adding it to openclaw.json, redeploying with stow, and confirming the gateway scheduler picked it up. ## Quick Start Ask the agent to add an OpenClaw cron job with a given job ID, agent ID, schedule type and value, and prompt, then deploy and verify it.