Back to browse
GitHub Repository

AVE - Agentic Vulnerability Enumeration. The behavioral classification standard for agentic AI components.

9 starsPython

Ave, a behavioral classification standard for agentic AI

by bawbel·Jul 21, 2026·4 points·1 comment

AI Analysis

●●●BangerBig BrainBold Bet

CVE-style stable IDs for AI agent exploits fills a critical standardization gap.

Strengths
  • Maps behavioral fingerprints directly to OWASP MCP Top 10 and MITRE ATLAS frameworks.
  • Schema includes AIVSS scoring system specifically designed for natural language exploits.
  • 100% test coverage on validation rules ensures strict schema adherence.
Weaknesses
  • Adoption depends entirely on security tooling vendors integrating the new schema.
  • Abstract behavioral definitions may be harder to automate than code-based CVEs.
Category
Target Audience

AI security researchers and compliance teams

Similar To

CVE · OWASP Top 10 · MITRE ATT&CK

Similar Projects

Weaver-spec – a shared contract so agent components compose

Specs for an agent ecosystem that doesn't have adoption yet.

Bold Bet
dgenio
2026d ago