promptdojo_promptdojo_promptdojo_promptdojo_promptdojo_promptdojo_promptdojo_promptdojo_promptdojo_promptdojo_promptdojo_

Add evals and traces — measure the agent, don't trust it — step 7 of 9

Cursor wrote this eval: it calls the agent, gets the answer, then asserts the answer equals what the agent returned. The eval ALWAYS passes — it's checking the agent against itself. Tautological.

Fix the eval to assert against an INDEPENDENT ground truth string ("Paris") that came from a fact-check, not from the agent.

Expected output:

pass: True
The break is on line 9 — but read the whole snippet first.

full-screen editor opens — close anytime to keep reading.