[GitHub Trending] Tencent/AI-Infra-Guard

7.5 relevance

AI red teaming platform for agent and MCP security, highly relevant to AI infrastructure and platform engineering.

AI/ML github.com

Answer the board on AI readiness

Arrive with evidence, not opinion. See how your foundation benchmarks against your peers, get a clear read, and leave with a ranked place to start.

Why It Matters

For platform and AI engineers building or securing LLM-based agents and MCP servers, this is the first open-source toolkit that unifies vulnerability scanning, jailbreak testing, and skill-level red-teaming into a single, actively maintained platform from a major AI lab.

Key Takeaways

Evaluate A.I.G v4.5.2 against your agentic AI stack to automate red-teaming for MCP tools, agent skills, and prompt injection vectors before production deployment.

A full-stack AI Red Teaming platform securing AI ecosystems via Agent Scan, Skills Scan, MCP scan, AI Infra scan and LLM jailbreak evaluation. - Tencent/AI-Infra-Guard

Summary

Tencent's Zhuque Lab open-sourced AI-Infra-Guard (A.I.G) v4.5.2, a comprehensive AI red-teaming platform integrating ClawScan, Agent Scan, MCP Server & Skills Scan, and Jailbreak Evaluation. The latest release adds .pyc bytecode bypass detection, charset smuggling defenses, RCE prevention via tool whitelisting in MCP-Scan, and a SkillJack research project, expanding the vulnerability library to over 2,000 CVE rules. Deployable via Docker with a web UI at localhost:8088, it supports standalone CLI tools for skill, MCP, and agent scanning, but lacks authentication and is intended for internal enterprise use only.

Author

Tencent