Beyond Task Completion: A Verification-vs.-Conformance Gap in Tool-Evolving Agents
Agents that synthesize their own tools ship a second artifact alongside each answer: a software library that future tasks reuse, compose, and depend on. Task completion (TC) certifies the answer; it does not certify the library. On a Claude Haiku 4.5 pilot, we patch the harness to preserve per-tool source and replay every synthesised tool against a held-out conformance suite. Across 222 preserved tools and three protocols, 96.8% record per-tool correctness C=0.00: two protocols silent-rot at 100%, one at 91.7%. Hand-written reference implementations score C=1.00 on all 16 capability suites. The synthesised tools execute cleanly and the in-session verifier raises no flag; on held-out inputs they simply return wrong answers. This is the verification-vs.-conformance gap a pass-rate benchmark cannot see. EvolveTool-Bench makes that gap measurable. Sessions are structured into seed, gap, variant, composition, regression, and adversarial roles; runs emit per-tool manifests, verified-subset TC, correct-vs.-incorrect reuse, and audit traces for post-deployment monitoring. In the Haiku pilot over five protocols (3 seeds, 8 sessions, 51 verified decisions per pass), TC alone does not separate the protocols; the audit layer does. One pre-specified contrast shows the adapted decision-split synthesis protocol underperforming one-shot synthesis (-7.1 pp, 95% CI [-13.9,-0.1]; not significant after BH correction) -- a hypothesis-generating signal, not a confirmed effect. We release the harness, per-tool source preservation, conformance replay, and reproducibility manifests. An audit schema of this shape -- verifier coverage, reuse decomposition, held-out conformance -- is the minimum a deployed tool-evolving agent should emit between releases.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Advice Conformance Verification by Reinforcement Learning agents for Human-in-the-Loop
Human-in-the-loop (HiL) reinforcement learning is gaining traction in domains with large action and state spaces, and sparse rewards by allowing the agent to take advice from HiL. Beyond advice accommodation, a sequentia…
Decision Makingreinforcement-learningReinforcement LearningReinforcement Learning (RL)+1Probabilistic Conformance for Cyber-Physical Systems
In system analysis, conformance indicates that two systems simultaneously satisfy the same set of specifications of interest; thus, the results from analyzing one system automatically transfer to the other, or one system…
Model Predictive ControlSentinelAgent: Intent-Verified Delegation Chains for Securing Federal Multi-Agent AI Systems
When Agent A delegates to Agent B, which invokes Tool C on behalf of User X, no existing framework can answer: whose authorization chain led to this action, and where did it violate policy? This paper introduces Sentinel…
GoTube: Scalable Stochastic Verification of Continuous-Depth Models
We introduce a new stochastic verification algorithm that formally quantifies the behavioral robustness of any time-continuous process formulated as a continuous-depth model. Our algorithm solves a set of global optimiza…
global-optimizationConformance Testing for Stochastic Cyber-Physical Systems
Conformance is defined as a measure of distance between the behaviors of two dynamical systems. The notion of conformance can accelerate system design when models of varying fidelities are available on which analysis and…
Conformal Prediction