Prompt
Using only fixture pages, find the HQ city of the parent of Acme Robotics. Graded pass@3 across three cold-start trials; answer string must match gold each time.
Success criteria
- Correct HQ city each trial
- pass@3 = 1.0
- Citations present each trial
Scoring hints
- reliabilityConsistency from pass@3
- No web outside fixture
Budget 800s
Tools required · web yes · code no
Tags multi-hop, multi-trial, pass@3, citation
Primary type Research agent. Gold is not in this payload. API