Measuring Agents in Production
arXiv:2512. 04123v4 Announce Type: replace-cross Abstract: LLM-based agents already operate in production across many industries, yet we lack an understanding of what technical methods make deployments successful.
arXiv:2512. 04123v4 Announce Type: replace-cross Abstract: LLM-based agents already operate in production across many industries, yet we lack an understanding of what technical methods make deployments successful.
arXiv:2607. 06713v1 Announce Type: cross Abstract: Large language models are rapidly moving towards closing the development cycle, transitioning from simple assistive companions to autonomous contributors deeply embedded into collaborative development environments.
WatchPoint is a simulated‑user system that generates and runs diagnostic scripts against a live web application, producing structured observations to guide coding model retries. Unlike prior methods that rely on screenshots or non‑executable metrics, WatchPoint operates on Web‑Bench—a benchmark of 50 multi‑file web projects with 1,000 sequential tasks verified by deterministic end‑to‑end tests. It recovers 57.6% of diagnosed tasks, matching a human tester’s 54.5% recovery rate, and identifies when such simulated feedback is beneficial or should be withheld.
arXiv:2610.16776v1 Announce Type: new Abstract: Game-Creation-Agents (GCA) turns a natural-language request into a runnable Phaser 4 browser game. The pipeline first builds a multimodal game design d...
arXiv:2607. 14336v1 Announce Type: cross Abstract: Trustworthy deployment of LLM-based agents in software systems requires evaluating how they perform on application-specific workflows, with enough granularity to localize where they succeed and fail.
arXiv:2607. 06624v1 Announce Type: new Abstract: We present AgentLens, a production-assessed benchmark for interactive code agents.
Consort is a spec‑first, test‑driven agent framework that enforces engineering discipline through immutable controls, a deterministic orchestrator, and human‑approved gates. It operates on live database branches, guiding role agents through a spec‑first design lane and a test‑driven build lane. The framework claims that enforcing tests and gates in code keeps agent‑written code honest and verifiable, while specialized roles make it maintainable.
AgentSWE is a new benchmark for agent software engineering that evaluates coding agents across three lifecycle stages: creating an agent from a natural‑language requirement, editing existing agent codebases, and optimizing toward a measurable target. The benchmark consists of 25 commission‑style tasks, where builders iterate on a few illustrative cases and are then tested on held‑out cases that cover the full requirement, with hard pass/fail gates or objective‑based grading. Experiments show that current coding agents fail to meet the requirements, revealing common failure modes such as overfitting, incomplete runtimes, and unsupported completion claims.
As Large Language Models (LLMs) evolve into autonomous agents, the need for unified evaluation infrastructure becomes critical. However, current evaluation pipelines remain highly fragmented and tightly coupled, hindering reproducibility and causing redundant engineering.
arXiv:2607. 19837v1 Announce Type: new Abstract: Traditional pentesting uses reconnaissance at each step to uncover unseen weaknesses, build stronger attacks, and advance the objective; we argue that AI agents require the same treatment.
arXiv:2608.29204v1 Announce Type: cross Abstract: Generative AI-based software engineering agents are becoming routine contributors to real-world software projects. On GitHub, developers can assign t...
CART (Closed‑Loop Adaptive Red Teaming) is a framework that iteratively uses results from red‑teaming tests to guide subsequent probes, thereby expanding risk coverage and maintaining diversity. It separates the roles of Challenger (test generator), Target (model or agent under test), and Judge (result evaluator), enabling independent study of each component. Across multiple evaluation families, CART uncovers more failures and higher risk than static prompt replay, demonstrating that adaptive, continuous testing reveals weaknesses that fixed‑prompt methods miss.