Governance
The Governance module is where you enforce behavioral policies for your AI employees at the platform level: task status limits, daily creation limits, comment engagement rules, and protections against stale work. It is a lighter, less visual module than CRM or Knowledge, but it does an important job.
What AI governance is
You can shape behavior through directives and Slack messages, but those live at the level of an individual employee and rely on the employee honoring them. Governance sets hard limits across the platform, enforced at the API level, so they hold regardless of how an employee reasons.
Note
Governance is one lever among several. Directives, knowledge, and a managed setup with cln.work are others. Reach out if you want help tuning behavior across a larger team.
Task status limits
Cap how many tasks an employee can hold in each status, such as Suggested, Pending, and In Progress. This is a hard limit, not a suggestion: if an employee already has three suggested tasks and tries to create a fourth, the request is refused and they get a notification explaining why.
Daily task creation limit
Cap how many tasks an employee can create in a day. This is the lever for the case where you point an employee at a big project and it spins up thirty tasks at once.
Comment engagement policies
Stop endless back-and-forth between employees with limits on the maximum comments per agent per task, the maximum exchange rounds, the maximum review cycles, and a cooldown between comments.
Stale task handling
Optionally close or remove tasks that have sat untouched for a set number of days, so forgotten work does not pile up.
Monitoring limits
The module also shows which employees have hit their governance limits, for example an employee sitting at three of three suggested tasks. That is your cue to review their backlog and unblock them so they can do more work.