ResearchAgentDropoutV2 Test-Time Rectify-or-Reject Pruning for Multi-Agent SystemsarXiv·high signalXBlueskyLinkedInCopy linkTackles cascading errors in multi-agent systems. Test-time pruning framework as active firewall between agent handoffs without retraining.SourceSource pagearXiv↳ Follow the threadPolicy dependency / Stack layerSCA-Agent reconstructs a dependency's trace across Code, Build, Release, Deploy and Runtime and beats the best traditional SCA tool by 18.76 points of F1arXiv 2609.18391Stack layer / Threat patternPentestChain keeps a 7B local model off the critical path behind a deterministic exploit map and runs an eleven-tool MCP pentest pipeline at zero paid-API costarXiv 2609.18120Stack layer / ContrastModel-Checking an Agent's Plan Before Any Tool Runs Rejects Unsafe Plans Without Spending a Single Tool CallarXiv 2609.18674Stack layer / ContrastThe Most Collusive Pricing Model Honestly Reports Cooperative Intent, So Chain-of-Thought Monitoring Cannot Catch ItarXiv 2609.18346Stack layerMulti-Agent Collaboration Pays Only on Long-Horizon Tasks With Sparse DependenciesarXiv 2609.19759Stack layerCairn makes agent memory collective: query the community's reputation for a tool before calling it, submit evidence afterarXivStack layerRunning an LLM Locally Doesn't Keep Prompts Private: They Survive in Allocator Memory After InferencearXiv 2609.18526Policy dependency / Stack layerA Cheap Read-Only Verifier Captures Nearly All the False-Pass Benefit of a Full Planning StackarXiv 2609.20474