A census of 86 self-described multi-agent repositories found that 68.2% are actually single-model or non-agent systems, according to a study published in SILICON SCIENCE. The analysis, which annotated…
A fine-tuned model called ReViSQL-K2.6, developed by Tinker using reinforcement learning with verifiable rewards, exceeds the human benchmark of 92.96% on the BIRD text-to-SQL task, reaching 92.96% wh…
A developer's comparison of AI agent frameworks in 2025 evaluates LangChain, CrewAI, and the broader ecosystem including MAF. The analysis highlights LangChain's comprehensiveness and LangGraph's suit…
Atoms, a new vibe coding tool built by the team behind MetaGPT, uses a multi-agent AI system to handle the entire product lifecycle—from market research and app development to deployment and marketing…
AgentNexus, a new open-source framework, coordinates multiple LLM agents by service boundaries rather than roles, enabling automatic diff-aware notifications when documents change. The system uses a v…
Agentic RAG systems that learn from past interactions are vulnerable to memory poisoning attacks, where attackers plant false memories that the system treats as ground truth. Researchers at the Univer…
Multi-agent AI frameworks like OpenHands and MetaGPT show failure rates above 85% in production conditions, with the root cause traced to data layer issues rather than model quality. A developer found…