Skip to main content
Manual inspection is the trace-level view for a single run. It is where you check the exact step that produced a surprising result. Sovara run inspection view showing a trace graph and selected step details Sovara run inspection view showing a trace graph and selected step details The left side shows the run graph. Each node is a model call, tool call, or recorded step. Selecting a node opens its input and output on the right, including injected priors, request metadata, and the content sent to the model or tool. Manual inspection helps with:
  • Finding the first step where the agent drifted
  • Checking which context the model actually saw
  • Reviewing tool inputs and outputs
  • Confirming whether a prior affected the run