Introduction Working with organisations that are attempting to create GitHub Copilot custom agents, take-up of these agents by their community becomes important to know. Some questions quickly emerge are "how well are we actually using it?", "which agents are getting used and which have not had that much traction?". Native metrics provide high-level insights into adoption, but they lack the depth needed to answer more granular questions—such as which agent workflows are most used, or how beha...

Read the original article