Replay & structural diff

Pick a replay mode. Read what each one guarantees. The structural diff below shows RUN_A vs RUN_B — same agent, different prompt version, different output.

Re-execute the agent code. Serve every model and tool call from the capsule cache.

Same control flow, same outputs, no network.

Mocked replay — RUN_A

replay_id: 01KT2P4Y8WGTHE9MRP4ZRX7QP3

✓openai.completion · served from cache (1240ms recorded → 0ms replay)
✓openai.completion · served from cache (980ms recorded → 0ms replay)
✓mcp.read-file · served from cache
✓mcp.run-tests · served from cache

2/2model calls mocked

2/2tool calls mocked

0env warnings

Structural diff

01KR5SQZPD… vs01KS9K8R2N…

2changed

0added

0removed

Environment0 changes

No environment changes.

Model calls1 change

2 aligned / 1 changed / 0 added / 0 removed

model_call[0]call_A_001 → call_B_001
no changes
model_call[1]call_A_002 → call_B_002
- request.messages[0].contentmajor
  −You are a thorough code reviewer. For the diff below, identify: (1) bugs, (2) edge cases the author may have missed, (3) style inconsistencies. Be specific. Quote line numbers.
  +Review this code diff. Note bugs and style issues.

Tool calls0 changes

No tool calls recorded.

Outputs1 change

outputs/review.md
before: a6911021ffa38800…after: 833f562a971d0ace…

On your machine

nova replay <run-id> --mode mocked
nova diff <run-a> <run-b>
nova diff <run-a> <run-b> --output github-annotation