# AVE MCP server

> Provides an open behavioral classification standard to identify, score, and map vulnerabilities in agentic AI components like MCP servers and system prompts.

- Repository: https://github.com/aveproject/ave
- Page: https://hostedmcp.org/servers/aveproject/ave
- Transport: —
- Stars: 26
- License: Apache-2.0

## FAQ

### What is AVE?

AVE is an open behavioral classification standard for agentic AI components, providing stable IDs, AIVSS scores, and behavioral fingerprints for potential weaponization vectors.

### Is AVE a security scanner?

No, AVE is a naming and classification standard. It defines behavioral vulnerability classes, while separate implementations like the bawbel-scanner perform the actual detection.

### How does AVE map to other security frameworks?

AVE maps its records to the OWASP MCP Top 10 and MITRE ATLAS frameworks so that findings can be integrated into existing security reporting workflows.