ZIRAN Agent Security Scan
Version updated for https://github.com/taoq-ai/ziran to version v0.4.0.
- This action is used across all versions by 1 repositories.
Action Type
This is a Composite action.
Go to the GitHub Marketplace to find the latest changes.
Action Summary
ZIRAN is a security testing tool designed to identify vulnerabilities in AI agents, including those with tools, memory, and multi-step reasoning, rather than focusing solely on LLMs or web applications. It automates tasks such as tool chain analysis, multi-phase attack simulation, and remote scanning of agents over various protocols (REST, OpenAI, A2A, etc.), enabling comprehensive assessments of an AI agent’s security posture. Key capabilities include knowledge graph tracking, protocol auto-detection, and support for CI/CD quality gates, making it a robust solution for safeguarding complex AI systems.
Release notes
What’s Changed
- feat: add OpenAIConfig model and handler forwarding by @leoneperdigao in https://github.com/taoq-ai/ziran/pull/6
- feat: implement BedrockAdapter for AWS Bedrock Agents by @leoneperdigao in https://github.com/taoq-ai/ziran/pull/7
- feat: add AgentCoreAdapter for Amazon Bedrock AgentCore by @leoneperdigao in https://github.com/taoq-ai/ziran/pull/8
- feat: add LLM backbone and LLM-as-a-judge detector by @leoneperdigao in https://github.com/taoq-ai/ziran/pull/9
- feat: add examples and dependency configuration by @leoneperdigao in https://github.com/taoq-ai/ziran/pull/10
Full Changelog: https://github.com/taoq-ai/ziran/compare/v0...v0.4.0