cs.CRcs.AI
The paper introduces a new benchmark and decomposition method, Sufficiency-Tightness Decomposition, demonstrating that current coding agents struggle to accurately infer least-privilege authorization, and that this decomposition significantly improves both security and task success.
As coding agents gain access to shells, repositories, and user files, least-privilege authorization becomes a prerequisite for safe deployment: an agent should receive enough authority to complete the task, without unnecessary authority that exposes sensitive surfaces. To study whether current models can infer this boundary themselves, we first introduce permission-boundary inference, where a model maps a task instruction and terminal environment to a file-level read/write/execute policy, and AuthBench, a benchmark of 120 realistic terminal tasks with human-reviewed permission labels and executable validators for utility and attack outcomes. AuthBench shows that authorization is not a simple conservative-versus-permissive calibration problem: frontier models often omit permissions required by the execution chain while also granting unused or sensitive accesses. Increasing inference-time reasoning does not resolve this mismatch. Instead, each model moves toward a model-specific authorization attractor: more reasoning makes it more consistent in its own failure mode, whether broad-but-exposed or tight-but-brittle. This suggests that direct policy generation is the bottleneck, because a single generation must both discover all necessary accesses and reject all unnecessary ones. We therefore propose Sufficiency-Tightness Decomposition, which first generates a coverage-oriented policy by forward-simulating the task and then audits each granted entry for grounding and sensitivity. Across tested models, this decomposition improves sensitive-task success by up to 15.8% on tightness-biased models while reducing attack success across all evaluated models.