A study of open-source LLM-based multi-agent systems identifies common issues, their causes, and potential solutions. The most common issue is orchestration and execution, with causes including workflow problems, tool integration issues, and memory problems. The study suggests optimizing workflows as a solution.

This paper introduces Revision-Aware Independent Agent Graphs (RIAG), a policy that enables dynamic task routing and manages document versions for AI agents. The authors evaluate RIAG on six benchmarks, achieving higher accuracy and lower call counts compared to existing methods.

A research paper introduces Controlled In-Context Memory (CICM), a benchmark for tracking and using updated information in conversations and agent logs. The study finds that even frontier reasoning models can fail to recover the current state, and attention drift is identified as a mechanism for this failure. The paper proposes a solution to correct old-value errors without retraining models.

A new training-free context compression framework, FOCUS, is introduced for LLM agents. FOCUS operates at test time and does not require offline data collection or fine-tuning, making it architecture-agnostic and easy to attach to closed-API frontier models. It achieves state-of-the-art performance on various agentic benchmarks, reducing context and dependency by up to 48% and 73%, respectively, while improving task success by up to 8.9 percentage points.

AnyAct proposes a universal action layer for self-evolving agents to tackle challenges in large-scale, dynamic tool ecosystems, including the scale dilemma, non-stationarity, and heterogeneity of feedback formats.

ThuRunel is a dynamic decoupling approach for structured advisory dialogue, combining a finite-state belief management framework, a chain-of-thought teacher synthesis protocol, and learned generation adapters. It has been deployed as a bilingual web application and achieved consistent improvements in elicitation completeness and specialist brief quality against eleven baselines.

This paper introduces Grid-Orch, an LLM-powered orchestrator for distribution grid simulation and analytics. It bridges LLMs and power system simulation through the Model Context Protocol, enabling engineers to perform complex analyses via natural language. The platform supports cloud-hosted and locally deployed models, includes 36 domain-specific tools, and extends to multi-step engineering workflows.

Cloudflare's Kitesurf browser for agents has been updated with support for WebMCP, improved WPT coverage, and efficiency optimizations, making it a more capable and efficient browser for AI agents.