AgentsAgentShield Benchmark: First Open Test of 6 Agent Security VendorsGitHub·high signalXBlueskyLinkedInCopy link537 test cases, 8 categories. Scores range 39-98. Critical: providers catching 95%+ prompt injections miss most unauthorized tool calls. Provenance verification nearly absent.SourceSource pageGitHub↳ Follow the threadPolicy dependency / Stack layeragent-hop Reverse-Engineers Five Agents' Session Formats So You Can Resume a Chat in a Different ToolGitHubStack layer / Threat patternNVIDIA's SkillSpector scanner hits v2.8.2 with a 42,447-skill study: 26.1% carry a vulnerability, 5.2% look outright maliciousGitHub (nvidia/skillspector)Stack layer / Threat patterncaveman Hits 97,000 Stars for a Token-Cutting Skill — and Its README Publishes the JetBrains Test That Drops the Claim From 65% to 8.5%GitHubStack layer / Threat patternECC v2.1 Adds a Browser-Based Plan Canvas and a Native Kimi Code Install PathGitHubStack layer / Threat patternopen-kritt Open-Sources Agent-Orchestrated Vulnerability Hunting on the Premise That Whole-Repo Scanning FailsGitHubStack layer / Threat patternNativ Turns a Mac Into an Anthropic- and OpenAI-Compatible Local Server With Prewired Configs for Ten Coding AgentsGitHubStack layer / Threat patternAgno patches an MCP flaw where a model could smuggle `tool_name` and execute any tool, bypassing every approval gateGitHub (agno-agi/agno)Stack layer / Threat patternagency-agents Hits 140,066 Stars With 230+ Agent Personas — and Ships a Native Installer App So You Never Clone the RepoGitHub Trending