What problem does it solve? Cron jobs in Hermes Agent silently fail for non-obvious reasons: the gateway process was not running, output was routed to an undeliverable WebUI session, or the global model/provider config drifted after the job was created. This Skill provides systematic diagnostic flows to identify the root cause and apply the correct fix. ## Core Features & Use Cases - Missed Run Diagnosis: Trace why overnight cron jobs never fired by checking the gateway process, agent logs, and the conversation-loop-dependent scheduler model. - Delivery Failure Resolution: Fix jobs that ran successfully but never reached the user, including the WebUI "unknown platform" mismatch and CLI/TUI deliver='local' pitfalls. - Provider/Model Drift Repair: Pin provider and model on unpinned jobs by directly editing ~/.hermes/cron/jobs.json when the CLI lacks edit flags. - Use Case: A user wakes up to find their nightly summary cron job never arrived. Use this Skill to check cronjob action='list', inspect last_delivery_error, discover the job targeted the WebUI, and update it to deliver='all' so results reach Discord or Telegram. ## Quick Start Diagnose why my scheduled cron job did not deliver its output last night and fix the delivery target.
