Agent run: title

• Date:
• Task class: API/configuration / pipeline/storage / migration/simulation
• Starting commit:
• Agent/tool version:
• Human-provided context:

Outcome

• Status: completed / incomplete
• Observable result:
• Files changed:
• Regression count and affected contract:

Cost

• Elapsed time:
• Tool/model cost, when available:
• Human interventions and reason:

Verification

• <command>: result and artifact location

Failure classification

Choose one: none / setup / navigation / requirements / implementation / verification / permissions.

• Evidence:
• Repository or tooling improvement suggested:

Review

• Reviewer:
• Reproduction result:
• Conclusion: