GitHub Repos· Oct 25, 2026· by Wapiti08

Wapiti08/AgentCrucible

Attack, observe, and secure agentic AI and MCP systems.

🎯 Vulnerability: Web Security Vulnerability

💡 Defense Fix: Implement strict input validation, contextual output encoding, and enforce least-privilege authorization controls.

#agentic-ai#ai-security#llm-security#mcp-security#runtime-security#sandbox#software-supply-chain#vulnerability-research