UNPKG

grix-connector

Version:

Connect local AI coding agents (Claude, Codex, Gemini, Qwen, DeepSeek, Cursor, OpenCode, Pi, OpenHuman, Reasonix) to the Grix scheduling platform. Also serves as an OpenClaw plugin for Grix channel transport.

57 lines (42 loc) 2.08 kB
# Grix Update Cron Setup This document defines the automatic maintenance integration after plugin installation is complete. ## Recommended Goal After installation is complete, there should be a scheduled cron task that periodically executes `grix-update`. Recommended requirements: 1. Use `isolated` session 2. Execute silently by default 3. Executed by the main agent responsible for maintaining this machine 4. Run once every 6 hours by default ## Recommended Create Command ```bash openclaw cron add \ --name "grix auto update" \ --every "6h" \ --agent <main_agent> \ --session isolated \ --light-context \ --no-deliver \ --message 'Use the grix-update skill with {"mode":"check-and-apply","plugin_id":"grix","notify_on":"never","allow_restart":true}. If there is no update or the update succeeds, reply exactly NO_REPLY. If the install is unsupported or any step fails, return one short failure summary.' ``` ## Handling Duplicate Installations If a task with the same name already exists, do not create a duplicate; use `edit` instead: ```bash openclaw cron edit <job_id> \ --every "6h" \ --agent <main_agent> \ --session isolated \ --light-context \ --no-deliver \ --message 'Use the grix-update skill with {"mode":"check-and-apply","plugin_id":"grix","notify_on":"never","allow_restart":true}. If there is no update or the update succeeds, reply exactly NO_REPLY. If the install is unsupported or any step fails, return one short failure summary.' ``` ## Verification Commands ```bash openclaw cron list openclaw cron status openclaw cron runs --id <job_id> --limit 20 ``` ## Notes 1. `--session isolated`: Avoid updates and gateway restarts interrupting the current main chat. 2. `--no-deliver`: Do not send messages externally on success. 3. `notify_on=never`: In cron scenarios, do not let the skill send notifications on its own; let the cron run records handle convergence uniformly. 4. If failure alerting is needed later, it should be handled in the upper-level cron rules; do not embed notification logic back into the skill.