AI Agent Studio: Configure Evaluation Run - Agent Version "Undefined" Producing Error
Summary:
When configuring an Evaluation Run in AI Agent Studio (Monitoring and Evaluation > Manage Evaluations) for the "Supply Chain Planning Constraints" agent, the Version dropdown lists two entries but both display the literal text "undefined" instead of an actual version identifier. Selecting either option clears the field, and clicking Run then fails with "Can not execute the evaluation run." This reproduces across multiple Evaluation records tied to the same agent.
Why this is blocking us: This agent has both a Draft and a Published version present in AI Agent Studio (currently Published v36, with an active Draft). The Version dropdown should list these two real, distinct versions — instead both render as "undefined," suggesting the version records exist server-side but their display-label field isn't populated for this agent.
