Probe AI agents for dangerous tool chains and execution side effects with ZIRAN

Test agent systems for graph-discovered tool-chain vulnerabilities, silent side effects, and multi-phase trust exploits before deployment.

Prerequisites

AI agent system or test target, supported adapter or protocol path such as MCP or A2A, Python environment, security test configuration

Installation

Use the upstream install or setup path that matches your environment:

• pip install ziran[ui]
• docker compose up

Requirements and caveats from upstream:

• 
• Or with Docker:
• image: python:3.12-slim

Basic usage or getting-started notes:

• <a href="#quick-start"><b>Quick Start</b></a>  · 

• <img src="docs/assets/adaptive-light.svg" alt="A live knowledge graph grows phase by phase: Reconnaissance discovers 3 capabilities, Capability Map adds 3 tools, Vulnerability Discovery surfaces a critical read_file→h...

• 

• Source: https://github.com/taoq-ai/ziran

• Extracted from upstream docs: https://raw.githubusercontent.com/taoq-ai/ziran/HEAD/README.md

Documentation

• https://github.com/taoq-ai/ziran#readme

Source

• Agent Skill Exchange