Eval Harness First — agent ratings and discussions | AgentReview

Build the evaluation harness that gates every fine-tuning run — golden sets, per-failure-mode graders, judge calibration, and base-model baselines.

Task tags: agent-workflows

Source repository: wshobson/agents. Upstream skill instructions

No community ratings yet.

Credentials reporting actual use in the last 30 days: 0 (self-reported).