SourcesIntentCUA Multi-Agent Desktop Automation 74.83% SuccessarXiv·high signalMulti-agent framework achieving 74.83% task success rate. Planner, Plan-Optimizer, Critic over shared memory. Cross-application skill transfer. Outperforms RL and trajectory baselines.SourceSource pagearXiv↳ Follow the threadStack layer / Update threadSLBench Finds Up to 70% Unsafe Execution When Codex and Claude Code Follow Logical Relations in Skill FilesarXivStack layer'Who Broke the System?' AgentLocate Pinpoints the Failing Agent and Step in Multi-Agent RunsarXivContrastShared Selective Persistent Memory Lifts Agentic Task Completion to 96% and Cuts Per-Call Token Cost 97xarXivPolicy dependency / Stack layerReason Less, Verify More: Deterministic Gates Catch a Silent Policy-Violation Failure in Tool-Using AgentsarXivStack layer / Threat patternMitigating Taint-Style MCP Server Vulnerabilities Without Touching Server Code (SPELLSMITH)arXivStack layer / Threat patternVaronis discloses 'Rogue Agent' — a Dialogflow CX flaw letting one permission poison a shared agent runtimeNxCodeStack layer / ContrastLong-Horizon-Terminal-Bench: Best Frontier Agent Clears Just 15% of Real Terminal Tasks, Mean 4.3%arXiv (Tencent Hunyuan) / HuggingFace Daily PapersSame source domain / Semantic neighborWebSwarm Uses Recursive Multi-Agent Orchestration for Deep-and-Wide Web SearcharXiv