According to the official announcement, Perplexity today open-sources Numbat, a detection and response layer designed for agent-centric scenarios, engineered to operate across diverse agent frameworks. Numbat primarily provides security teams with visibility into agent activities, enabling relevant teams to observe agent behaviors across different frameworks and preemptively block selected operations before execution. This project focuses on agent security defense and operational control, with core functionalities including activity detection, response orchestration, and pre-execution interception.
Disclaimer: Contains third-party opinions, does not constitute financial advice
Tencent Hunyuan team open-sources the speculative decoding framework AngelSpec, and releases drafter weights and training code for Hy3-A21B
2 hours ago
Stanford AI Lab Releases SWE-chat, a Real-User Coding Agent Interaction Dataset
4 days ago
LangChain April Update: LangSmith Evaluator Templates and Deep Agents Deployment Commands
4 days ago
Microsoft Research will release and demonstrate the new version of MSR AI Frontiers on May 14, 2026
4 days ago
Avoca completes oversubscribed $125M funding round, achieving a valuation of $1B, set to build AI agents for the service economy
4 days ago
Google Cloud Unveils GEAR Agent Learning Path and Teases AI Infrastructure and Certification Plans for Next '26
4 days ago
Google Gemma Launches In-Browser Native AI Agent Gemma 4 with Native Tool Calling and No Server Required
4 days ago







This column focuses on the real progress of Agents: technological evolution, application implementat
Spotlight on Frontier, trending projects, and breaking events
Plain-English guides to complex ideas—your blockchain starter from basics
Crypto-stock linkage, real-time market quotes and in-depth analysis
Selected potential airdrop opportunities to gain big with small investments