patil-sudu237 / migration-guard-evaleval
Measures whether an AI agent reliably identifies production risks in database migrations and recommends safe rollout strategies.
0
manifest.yaml381 B
metadata: name: migration-guard-eval id: migration-guard-eval description: Measures whether an AI agent reliably identifies production risks in database migrations and recommends safe rollout strategies. license: MIT version: "1.0.0" dataset: file_path: dataset.jsonl features: input: scenario expected: expected metrics: - name: Pass rate threshold: 0.8