Agent View
Agent View is the whole platform scoped to a single agent: its cost and performance, the context and tools it uses, every conversation it has had, and more. The header shows the agent's current version, whether it's Running or Stopped, and when it was last updated. From there you can Test the agent in a conversation, Edit its configuration, Stop it, compare versions, or pin it to the sidebar.
Monitoring the agent
The Overview tab shows how the agent is doing over the selected date range, with each metric compared against the previous period. Sessions, traces, observations, and scores mean the same as on Home, scoped to this agent. Two metrics are specific to Agent View:
Avg. traces per session: how many requests a conversation takes, on average. A rise can mean customers need more back-and-forth to get what they came for.
Avg. iterations per trace: how many reasoning steps the agent takes per request. A rise can mean policies or routines are making each decision harder.

Below the metrics, Error Summary shows where errors concentrate: the most frequent error types, and the traces and sessions with the most errors. It's the starting point for any investigation.
Usage shows what the agent relies on most: the documents it queried, the policies and routines it matched, and the tools it called. Use Show lowest to reverse the order and find content the agent rarely or never uses, which is usually a candidate for review.
Recent Activity lists the latest deployments and context updates, with who made each change and when. When behavior shifts suddenly, check here first for what changed.
What the agent uses
The Context tab lists the content the agent pins, by type: Policies, Routines, Macros, Variables, Prompts, Glossaries, and Collections. Content is shared across the project, so each item also shows the other agents that use it. Check this before changing anything that more than one agent depends on.

The Tools tab lists the MCP servers connected to the agent, custom or built in, and the tools each one provides. Use Test on any tool to call it directly, which helps tell a tool problem apart from an agent problem.

Improving the agent
The Improvement tab collects the agent's Annotations, Experiments, and Scores. Items appear here when they are tagged with the agent, as agent:<agent-name>.
Changing the agent
Edit opens the agent's configuration, with its revisions listed on the side. Edit through the Form or directly in YAML. The form groups the configuration into Identity, System prompt, Context, Extra rules, Connections, Advanced, Additional fields, and Deployment.
When the agent is running, you are editing the live agent:
Save draft keeps your changes as a pending version without affecting the running agent.
Deploy configuration applies them immediately and creates a new revision.
Before deploying, use Test to try the changes in a conversation. Use Compare Revisions to see exactly what differs between two revisions: changed sections are expanded and marked Modified, and unchanged sections are collapsed.
Ask the Copilot
"Why did this agent choose that routine step in session 438844?"
Reads the session's traces and explains the decision.
"Show today's sessions with errors for this agent."
Returns the sessions, filtered by status.
"Which policies has this agent not matched in the last month?"
Ranks the agent's policies by usage.
"Is this agent stuck deploying?"
Checks the agent's logs and startup health.
"Roll back this agent to the previous revision."
Prepares the rollback and waits for your approval.
Last updated
Was this helpful?

