llm-security-benchmark

Star

Here are 5 public repositories matching this topic...

mega-edo / mega-security-leaderboard

Star

Leaderboard Comparing LLM Agent Security on System Prompt Leakage and Attack Probes

llm-security prompt-security agent-security agent-security-benchmark llm-security-benchmark prompt-security-benchmark

Updated May 2, 2026
Python

Ujwal-Ramachandran / AutoRedTeam_LLM_Pipelines

Star

Automated red-teaming pipeline benchmarking LLM safety across 150 adversarial prompts. Achieves 81.1% ASR across 3 models with zero human annotation.

python jailbreak adversarial-machine-learning red-teaming huggingface streamlit llm-security llm-security-benchmark

Updated Jun 10, 2026
Python

ScottColeSW / Agent-Defense-Theater

Star

A governed range for testing AI agent defenses: Agent Theater’s earned-trust tiers and human oversight, plus an adversarial arena that attacks each OWASP Top 10 for Agentic Applications risk. Measured results, honest gaps, and what it would take to close them.

human-in-the-loop red-teaming mitre-atlas llm-security owasp-llm-top-10 ai-agent-security agent-governance llm-security-compliance-prompt-injection llm-security-prompt-injection llm-security-probe llm-security-benchmark owasp-agentic-top-10 llm-security-prompt-injection-ai-safety llm-security-proxy llm-security-testing llm-security-audit llm-security-tools

Updated Oct 17, 2026
Python

Hcxgraphics / JailBreak-AI

Star

Modular LLM security research framework for benchmarking prompt jailbreak attacks, defense strategies, and LLM-as-Judge evaluation using Python, Ollama, FastAPI, and Mistral 7B.

nlp jailbreak python3 cybersecurity vulnerabilities red-teaming fastapi llm llm-security mistral-7b ollama-api llm-security-benchmark llm-security-prompt-injection-ai-safety

Updated Jun 27, 2026
Python

Bhardwajtanuj / Redteam-evaluator-LLM-StressTest

Star

Sandboxed red-teaming harness for LLM agents: an attacker agent tries prompt injection, indirect injection and jailbreaks on a defender agent, and a judge scores attack success rate, policy violations, trajectory errors and cost.

python jailbreak ai-safety ai-agents red-teaming prompt-injection llm-security llm-evaluation agent-evaluation prompt-injection-defense adversarial-testing prompt-injection-detection llm-security-benchmark llm-security-prompt-injection-ai-safety llm-security-testing

Updated Oct 16, 2026
Python

Add this topic to your repo

To associate your repository with the llm-security-benchmark topic, visit your repo's landing page and select "manage topics."

Learn more