The paper introduces Harness-Bench, a diagnostic benchmark that measures how different system 'harnesses' affect LLM agent performance in realistic workflows, showing that agent capability must be rep…

HarnessForge introduces a meta-adaptive framework that jointly evolves the execution structure (harness) and the reasoning policy of LLM agents, significantly improving overall system performance acro…

The paper distinguishes between a model's ability to generate useful updates for external agent components (harness-updating) and its ability to benefit from those updates (harness-benefit), finding t…

Adaptive Auto-Harness introduces a framework that enables LLM agents to sustain self-improvement and maintain high performance over open-ended, shifting task streams, outperforming existing fixed-benc…

This paper presents an approach for turning exploratory large language model prototypes into auditable applications with traceable, auditable architecture.

The paper proposes HiSME, a lightweight hierarchical skill meta-evolving solution that jointly optimizes skills and the skill evolving strategy by learning meta-skills from task execution traces, lead…

The paper introduces BenchTrace, a novel benchmark designed to rigorously evaluate the self-evolution and reflection capabilities of LLM agents, revealing that current models struggle with accurate fa…

The paper introduces AgentFlow, a novel framework that uses a typed graph DSL and feedback-driven optimization to automatically synthesize and improve multi-agent harnesses for discovering security vu…

The paper introduces Cochise, a minimal, reusable Python reference harness designed to standardize and simplify autonomous penetration testing experiments for comparing different LLM-driven agent arch…

The paper introduces Unified Context Evolution (UCE), a gradient-free framework that externalizes and manages agent experience into a typed, evolving library, significantly improving performance on mu…

OpenForgeRL is an open-source framework for training harness-based AI agents end-to-end in various environments using a lightweight proxy and Kubernetes orchestrator.

The paper presents a tool-making pipeline for production LLM agents that compiles repeated steps into validated, versioned tools before deployment, reducing latency and error rate.

The paper introduces a data-centric optimization pipeline to improve coding agents' ability to interact with a branching lakehouse, showing significant accuracy gains by treating agent evaluation as a…

This paper proposes a context-aware AI framework for generating telecom test scripts that adapt to fine-grained changes in the system using a delta engine and a knowledge graph.

The paper proposes Meta-Team, an experience-driven framework that enables multi-agent systems (MAS) to collaboratively self-evolve by transforming complex execution experiences into reusable improveme…

SkillAdaptor is a novel, training-free framework that enables stable, step-level adaptation of external skills for LLM agents by precisely attributing failures to specific skills.

The paper introduces Proteus, a self-evolving red-team framework that measures the adaptive leakage risk of LLM agent skills, demonstrating that current vetting methods significantly underestimate res…

The paper proposes TraceDev, a multi-agent framework for automated software development grounded in use cases, achieving higher success rates than baseline approaches in repository-level code generati…

The paper argues that current 'on-the-fly' AI agent design lacks necessary software engineering rigor and proposes an 'AI Workflow Store' to provide hardened, reusable, and reliable agent workflows.