eval
alirezarezvani1.4k installs
Find the best result from your AgentHub agents.
What it does
Ranks completed AgentHub agent results and identifies a winner using metrics, qualitative comparisons, or both.
When to use it
Use when the user invokes /hub:eval or asks to score, compare, or select among completed AgentHub agents.
How to use it
Give it an AgentHub session in the project; it returns ranked agent results with a highlighted winner.
What you provide
- An existing project
Latest review
What you need · 1
Requires Python to run the session-state update after evaluation; no version is specified.
Access · 0
No additional actions listed.