Summary
On August 7, 2026 GitHub expanded the Copilot usage metrics API to include agent application activity, giving organizations deeper visibility into automated task execution alongside human Copilot usage. The addition lets admins measure and govern agent-driven work through the same metrics pipeline.
What changed
The GitHub Copilot usage metrics API now reports agent app activity, extending its coverage from interactive Copilot usage to automated agent execution so admins can track agent-driven work programmatically.
Why it matters
As coding agents take on more autonomous work, organizations need to measure what those agents actually do, not just what developers type. Surfacing agent activity through the metrics API gives platform and engineering-leadership teams the data to attribute output, manage spend, and set policy on autonomous agents the same way they already track human Copilot adoption.
Evidence excerpt
Copilot usage metrics API adds agent app activity, providing deeper visibility into automated task execution.