Evaluation result artifacts for HiveSpec, tested with AgentV. Each run directory contains JSONL indexes, per-test grading, timing, and response artifacts in the standard AgentV output format.
| Plugin | Status | Results |
|---|---|---|
| hivespec | PR #824 | 17/17 Claude CLI, 9/17 Pi CLI |
Results can be reproduced by running the eval suite in the agentv repo:
cd agentv
agentv eval run --target <target> "evals/hivespec/*.eval.yaml"