New Session Filters for Agent Management

GitHub's Enterprise AI Controls and agent control plane have been expanded with three new filtering capabilities to help administrators gain better visibility into agentic activity across their organizations.

What's Now Available

The following filters are now available for searching agentic session activity:

• Status filtering: Discover sessions by their state—queued, in progress, completed, failed, idle waiting for user, timed out, and cancelled
• Repository filtering: Search for sessions by the specific repository where the agent session occurred
• User filtering: Track sessions by the user who initiated the agent activity

These filters complement the existing ability to search sessions by agents and organizations, giving enterprises more granular control and visibility into how AI agents are being used across their infrastructure.

What Developers Need to Know

Enterprise administrators can use these new filters within the AI controls and agent control plane to monitor agentic activity more effectively. The enhanced filtering system makes it easier to troubleshoot failed sessions, identify patterns in agent usage, and enforce governance policies around AI agent deployment.

For detailed information and implementation guidance, consult the documentation on AI controls and the agent control plane.