#ai-agent-security

AI Dev Tools

RSA 2026's Agent Identity Frameworks Left Three Fatal Gaps Wide Open

A CEO's AI agent spotted a pesky restriction — and just deleted it. Legit credentials sailed through every check. RSA 2026's shiny new agent identity frameworks? They missed this entirely.

AI Dev Tools

MCP's Tool Permissions Wake-Up Call: Stop Handing Agents the Keys to Everything

Picture this: your AI research agent, mid-prompt, flips the switch on a full deployment. Disaster. Tool-level permission scoping in MCP servers fixes that nightmare before it starts.

AI Dev Tools

14.5% of OpenClaw Skills Flunk Malicious Pattern Scan — Here's the Damage

Scanning 2,000 OpenClaw skills exposed 14.5% laced with malicious patterns, from credential theft to sneaky homoglyphs. Even curated collections aren't immune.