agentic-redteam is a security benchmark harness for frontier AI agents. It provides tools for red-teaming, jailbreak detection, and generating standardized reports (SARIF, OWASP aligned). It helps AI developers and security researchers evaluate and harden autonomous agents against adversarial inputs and vulnerabilities.
agentic-redteam is a LLM eval & observability project. Systematically testing and benchmarking the security, robustness, and vulnerability of AI agents to adversarial attacks and jailbreaks. It is built as an open-source project for AI security researchers. The project is open source (Apache-2.0). agentic-redteam is available on the command line.
It is developed by Muneeb, and it first shipped in 2026. The project is developed in the open on GitHub with 48 commits in the last 90 days. Among its 4 catalogued features are Red Teaming, Security Benchmarking, and Jailbreak Detection.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do