Skip to content

Reality Is the Final Verifier: On Two Key Gaps in Agentic Software Engineering

Sep 2026 · arXiv.org · Vol abs/2609.12039 · 2 citations · 123 references
Computer Science

TL;DR

An assurance-revision loop that uses deployment evidence to revise the requirements, model, or evaluator when stakeholders reject the resulting behavior is proposed, which cast assured agentic development as a resource-allocation problem over human judgment, agent capability, and compute.

Abstract

Software development follows an implementation-verification loop in which developers or agents iteratively revise an implementation until an evaluator, such as a test suite, accepts it. The evaluator checks the implementation against a set of requirements under a model of the deployment environment. Yet even a formal proof that the implementation satisfies the requirements under the model cannot guarantee acceptable behavior after deployment. Requirements only approximate stakeholder intent, and the model only approximates the real deployment environment. We call these together - requirement gap and model gap - the two-gap framework, which unifies the main failure modes of agentic software engineer-ing: reward hacking exploits omissions in the requirements or model, while hallucination widens the gaps by fabricating requirements or environment assumptions. Because neither gap can generally be certified closed in an open, changing world, the goal shifts from closing them to continuously narrowing them. We therefore propose an assurance-revision loop that uses deployment evidence to revise the requirements, model, or evaluator when stakeholders reject the resulting behavior. We then cast assured agentic development as a resource-allocation problem over human judgment, agent capability, and compute. The two principal bottlenecks mirror the two gaps: human judgment for the requirement gap and faithful, costly evaluation for the model gap. Reality remains the final verifier: acceptable behavior under actual deployment conditions is the ultimate test, while predeployment evaluations remain proxies for it.

View source

Similar papers

Preprint Oct 2026

Who Verifies the Builder? Toward Independent Assurance for Agent-Produced Software

Coding agents can now implement a specification, write and run tests, repair failures, and declare the result ready for release. When the same agentic system produces both the software and the evidence of its correctness, a passing test suite is no longer independent evidence: visible tests become an optimization targe...

Sion Israel Sion, Alain April · 0 citations
Book Open access Oct 2026

Extender: A Multi-agent System for Code Extensibility Review

Software systems must evolve as requirements and environments change, which makes extensibility a central quality goal. The SOLID principles, five object-oriented design principles, are widely endorsed in software engineering practice to evaluate software quality. However, automatically detecting violations of SOLID pr...

Jia-He Chen, Jia-Cheng Shen · 0 citations

From Capability to Assurance in Autonomous Penetration-Testing Harnesses: A Framework and Reference Implementation

This paper defines a framework of five assurance properties and claims an initial existence argument that the properties are realizable together, not a comparative performance result, and specifies the multi target, ablation, and adversarial evaluation protocol required to turn the framework obligations into measuremen...

Joas Antonio dos Santos Barbosa · 1 citation
Sep 2026

Verification as an Architectural Layer for LLM Agents: A V-Model Design, and a Pilot Study of Its Deterministic Core

Large language model (LLM) agents built on the ReAct pattern concentrate four responsibilities in one model: selecting a strategy, choosing each action, formatting it, and judging whether the result is adequate. Nothing outside the generative loop can reject its output, so an agent that cannot progress does not report...

Ali Afoud, J. Wu · 0 citations
Preprint Oct 2026

AgentProofBench: Benchmarking Reusable Verifiers for Web Agents

This work introduces AgentProofBench, which connects paired tests of a frozen acceptance rule to execution from identical saved openings, and makes reusable verification assessable at both the requirement and policy levels.

Jing-Jie Ning, Xue-Qi Li, Yi Lu et al. · 0 citations
Book Open access 2027

From Specification to Kernel Commit: Verified Code Generation on Real-World Systems

Verified code generation, in which an LLM produces both an implementation and a correctness proof, offers a promising path toward correct-by-construction software. Prior evaluations report that LLM capability remains unsatisfactory, but they test models in a bare-model setting that no longer reflects how LLMs are deplo...

Hao Sun, Ze-Nan Li, Cong Li et al. · 0 citations

Related blog posts

Microsoft Research Blog Nov 12, 2026

Microsoft Research Focus – Edition #98

What's new in research at Microsoft? Host Anson Ho explores the latest, including AI-powered universal translation, a breakthrough in a 31-year-old mathematics challenge, and advances making AI coding agents faster and more efficient.  The post Microsoft Research Focus – Edition #98 appeared first on Microsoft Research.

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.