Search

Check what an LLM is given for prompt injection: material at ingest, tool output, and the turn it answers

Adversarial fuzzer for MCP servers — dynamically break Model Context Protocol endpoints. Shell injection, SSRF, overflow, type confusion, prompt injection probes. Zero-dep Python core.

PolicyShield plugin for OpenClaw — runtime tool call policy enforcement

Open-source AI agent firewall. CEL policies, runtime enforcement, and compliance for LLM agents.

Make consequential AI-agent actions safe to execute.

CyberAI — AI-native multi-agent pentest platform

Nx plugin for [agent skills](https://agentskills.io/specification): any directory containing `SKILL.md` becomes a project with a full skill lifecycle — `build`, `lint`, `validate`, `os-check`, `size-check`. No `project.json` needed.

Alpha agent trust middleware: audit trails, safety guardrails, HITL, and tool validation

AI Bill of Materials — discover and inventory all AI/LLM agents, models, and API integrations across your infrastructure.

Project AIR: forensic reconstruction and incident response for AI agents. Turn agent traces into signed forensic records with BLAKE3 + Ed25519.

One-line safety middleware for AI agent APIs. Prompt injection scanning, cost budgets, decision audit trails. Works with FastAPI, Flask, and any ASGI/WSGI framework.

Universal cyber assistant: ~80 audited Kali tools + an expert skills library, driven by any model over MCP or direct run

A fail-closed gateway between an AI agent and its tools: normalize, default-deny policy, human approval, append-only audit.

TypeScript SDK for Lelu — the authorization engine for autonomous AI agents

Autonomous Penetration Testing AI Agent

AI model: hotdogs/Agents-A1-4B-cyber-lora (text-generation)

Contained static file analysis with Claude Code and Codex: REA, Ghidra and JADX in restricted Docker containers, selected input copies, readable reports. Linux/macOS; Windows experimental.

Senior Security Auditor AI agent — performs thorough read-only security analysis of codebases. Identifies vulnerabilities, assesses risk via CVSS scoring, maps compliance gaps, and provides actionable remediation. Built with LangChain, LangGraph, and Groq

AgentSec Inspector — AI Red Team and security auditing framework for LLM-powered agents, featuring direct and indirect prompt injection testing, MCP server and tool enumeration, MCP security auditing, RAG pipeline security assessments, and jailbreak resistance testing.

AEAP Composition Boundary Adapter for Google ADK. Thin wrapper that attaches the Runtime Decision Point to agent tool invocations.