Automated Software Testing

While recent LLM-based agents can identify many candidate bugs in source code, their reports remain static hypotheses that require manual validation, limiting the practicality of automated bug detection. We frame this challenge as a test generation task: given a candidate report, synthesizing an executable proof-of-concept (PoC) - such as a script, command sequence, or crafted input - to trigger the suspected defect. Automated PoC generation can act as a scalable validation oracle, enabling end-to-end autonomous bug detection by providing concrete execution evidence. However, naive LLM agents are unreliable validators: they are biased toward "success" and may reward-hack by producing plausible but non-functional PoCs or even hallucinated traces. To address this, we present ANYPoC, a general multi-agent framework that (1) analyzes and fact-checks a candidate bug report, (2) iteratively synthesizes and executes a PoC while collecting execution traces, and (3) independently re-executes and scrutinizes the PoC to mitigate hallucination and reward hacking. In addition, ANYPoC also continuously extracts and evolves a PoC knowledge base to handle heterogeneous tasks. ANYPoC operates on candidate bug reports regardless of their source and can be paired with different bug reporters. To demonstrate practicality and generality, we apply ANYPoC, together with a simple agentic bug reporter, on 12 large-scale, critical software systems, including Firefox, Chromium, LLVM, OpenSSL, SQLite, FFmpeg, and Redis. Compared to the state-of-the-art coding agents, e.g., Claude Code and Codex, ANYPoC produces 37% more valid PoCs for true-positive bug reports and rejects 9.7x more false-positive bug reports. ANYPoC also enables the discovery of 121 new bugs from over two thousand noisy bug reports, with 108 confirmed by developers and 92 fixed. 46 PoCs have also been adopted as official regression tests.

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.

Testing RESTful API is increasingly complicated but indispensable to quality assurance of cloud-native applications. This paper reports a multi-agent system called MASTEST that combines LLM-based intelligent agents and programmed agents to automate REST API testing. They form a complete tool chain covering the whole workflow of REST API test with API specification in the OpenAPI Swagger format as the input. It also incorporates human testers in the process to review and correct LLM generated test artefacts to control the quality of testing activities. MASTEST is evaluated on two LLMs, GPT-4o and DeepSeek V3.1 Reasoner with five public APIs. Its performances on various testing activities are measured by a wide range of metrics, including adequacy and coverage metrics, the syntax and data type correctness of generated test scripts, the usability of LLM generated test cases and scripts, as well as the bug detection ability. Experiment results demonstrated that both DeepSeek and GPT-4o achieved a high overall performance but had strengths and weaknesses on different testing activities. MASTEST generated test cases achieved 94% and 98% unit test coverage and 79% and 78% system test coverage for GPT-4o and DeepSeek respectively in comparison with human designed test cases. The generated test scripts maintained 100% syntax correctness and only required minimal manual edits for semantic correctness. The generated test scripts contain assertions on the expected status code as well as contents in the response messages. They are highly capable of detecting bugs in the REST APIs. Experiment data shows that the bug detection rates are between 2.13 to 4.50 per operation. These findings indicate that MASTEST is highly efficient and effective.

Scenario-based testing with driving simulators is extensively used to identify failing conditions of automated driving assistance systems (ADAS). However, existing studies have shown that repeated test execution in the same as well as in distinct simulators can yield different outcomes, which can be attributed to sources of flakiness or different implementations of the physics. In this paper, we present MultiSim, a novel approach to multi-simulation ADAS testing based on a search-based testing approach that leverages an ensemble of simulators to identify failure-inducing, simulator-agnostic test scenarios. During the search, each scenario is evaluated jointly on multiple simulators. Scenarios that produce consistent results across simulators are prioritized for further exploration, while those that fail on only a subset of simulators are given less priority, as they may reflect simulator-specific issues rather than generalizable failures. Our empirical study, which involves testing three lane-keeping ADAS on different pairs of three widely used simulators, demonstrates that MultiSim outperforms single-simulator testing by achieving, on average, a higher rate of simulator-agnostic failures by 66%. Compared to a state-of-the-art multi-simulator approach that combines the outcome of independent test generation campaigns obtained in different simulators, MultiSim identifies, on average, up to 3.4X more simulator-agnostic failing tests and higher failure rates. To avoid the costly execution of test inputs on which simulators disagree, we propose to predict simulator disagreements and bypass test executions. Our results show that utilizing a surrogate model during the search retains the average number of valid failures and also improves efficiency. Our findings indicate that combining an ensemble of simulators is a promising approach for the automated cross-replication in ADAS testing.

We consider the automatic online synthesis of black-box test cases from functional requirements specified as automata for reactive implementations. The goal of the tester is to reach some given state, so as to satisfy a coverage criterion, while monitoring the violation of the requirements. We develop an approach based on Monte Carlo Tree Search, which is a classical technique in reinforcement learning for efficiently selecting promising inputs. Seeing the automata requirements as a game between the implementation and the tester, we develop a heuristic by biasing the search towards inputs that are promising in this game. We experimentally show that our heuristic accelerates the convergence of the Monte Carlo Tree Search algorithm, thus improving the performance of testing.

Generated operational programs are often validated with either a few hand-written examples or exhaustive regression suites. The former can miss sparse boundary and interaction faults, while the latter can be unnecessarily expensive. We introduce FaultLens, a method for learning compact behavioral test suites while preserving an auditable connection to executed evidence. It executes a rich probe domain once, stores the fault-probe kill relation as a sparse outcome cache, and learns probe orderings only from earlier program generations. A fault-driven greedy component exploits known kill structure, while a mutation-independent diversity component covers probe families, cases, templates, and temporal bins. Their alternating hybrid remains useful when a new program contains a fault mechanism absent from ordering construction. We evaluate twenty generated operational policies across four environments, ten execution seeds, 1,200 measured run summaries, 2,160 controlled program transformations, and 4,120,200 executed program-probe pairs. Of 1,960 intended faulty transformations, 1,779 alter a contract or output somewhere in the finite audit domain; 200 additional controls preserve behavior. A 32-probe hybrid learned on generations 1-3 covers 576/582 (99.0%) dynamically killable faults in generations 4-5 using 1.2-2.0% of the exhaustive domain. With an entire fault family withheld from training, diversity raises scenario-family macro coverage from 84.6% to 94.9%. In a downstream deployment study, a conservative admission rule reduces severe tail regressions from 15/20 program-environment groups to 0/20. FaultLens provides a prioritized evidence mechanism, not a proof of correctness, and makes its budget, evidence source, generalization split, and misses explicit.