Skip to main content
The Coding Agent Governance overview pulls counts and trends from every event your hooks have captured. Use it to answer the question, “How is Coding Agent Governance going across my organization right now?”

Open the overview

Select Agent Governance from the left sidebar. The overview opens by default. Use Time Range at the top to scope every metric and chart to a window, such as the last hour, day, week, or a custom range. Coding Agent Governance overview

Top metrics

Five counts summarize fleet-wide activity. Hover any metric to see its definition.

Composition charts

Three pie charts show what your fleet is using during the selected time range.
  • Top Models: The top three models by session count.
  • Top MCP Servers: The top three MCP servers by tool-call count.
  • Skills: Loaded skills bucketed as Unauthorized, Suspicious, Approved, or Unscored by their Endor Score.
Use these to spot a model, an MCP server, or a skill gaining ground that you have not approved.

Top violation widgets

Below the pies, three card tables surface the busiest policy violations in the selected time range. Each card shows a few rows and links to Policy Violations with the matching filter applied.
  • Most Triggered Policy Violations: Top policies by violation count, across every category.
  • File Access Blocked: Top File Access matches that the hook stopped.
  • Dangerous Commands Blocked: Top Dangerous Command matches that the hook stopped.

When the overview is empty

The overview shows No data available when nothing matches the current time range. Confirm:
  • Time Range is wide enough to include relevant events.
  • Hooks are deployed and reporting. See Review your agent inventory to confirm activity is arriving.

Next steps

Continue with the following pages: