b456845e85
Adds promptfoo eval harness for agent quality scoring. LLM-as-judge system scoring task completion, instruction adherence, identity consistency, deliverable quality, and safety. Includes tests.
1.1 KiB
1.1 KiB