Long-horizon coding agents require timely corrections, but feedback can be ineffective or even harmful when it misjudges ongoing work or fails to address the underlying problem. According to the paper Opera: A Verbal Critic Framework for Long-horizon Coding Agents, existing critics focus on evaluating trajectories and generating feedback but rarely track what happens after that feedback is delivered.
Opera addresses this gap by treating each correction as a persistent note that is followed until the diagnosed problem is resolved. The framework decides when to review through periodic and event-driven triggers, diagnoses issues using typed operators, audits feedback against visible evidence before delivery, and tracks the agent's subsequent actions to distinguish mere compliance from actual resolution.
As a test-time critic, Opera improves the resolve rate of non-critic agents by up to 12.4, 15.0, and 8.9 percentage points on Terminal-Bench 2.1, a SWE-Bench Pro subset, and DeepSWE v1.1, respectively, across four policy models. It also achieves the highest mean resolve rate among competitive critic baselines on all three benchmarks, and improves policy models when the policy critiques itself.
Beyond inference, Opera-guided rollouts provide approximately on-policy training data. Fine-tuning Qwen3.5-9B on these rollouts improves its resolve rate on held-out SWE-Bench Pro repositories by 10.2 percentage points without a critic at inference time, matching fine-tuning on rollouts from a stronger model, while preserving performance when switching harness from OpenHands to Terminus-2 — a switch that substantially degrades the stronger-model approach.
The authors state that code is available at https://github.com/dongyuanjushi/Opera.