SkillFortify
LIVEAgent Skill Security Analysis
SkillFortify is the first formal security analysis tool for AI agent skills. Using the DY-Skill Dolev-Yao formal model, it detects skill injection, prompt leakage, unsafe tool invocations, and supply chain attacks across 22 agent frameworks — with zero false positives.
Features
100% Precision
Zero false positives. Every vulnerability flagged is a real vulnerability. Formal verification guarantees.
22 Framework Support
LangChain, CrewAI, AutoGen, OpenAI, Claude, Semantic Kernel, and 16 more agent frameworks.
DY-Skill Formal Model
Dolev-Yao model adapted for agent skills. Sound static analysis with formal security proofs.
ASBOM Generation
Agent Software Bill of Materials. Track every skill, its dependencies, and their security posture.
540-Skill Benchmark
Validated against 540 real-world agent skills. 96.95% F1 score.
3 Academic Citations
Already being cited by other security researchers. Recognized in the academic community.
Use Cases
Research Paper
SkillFortify: Securing the AI Agent Skill Supply Chain
Varun Pratap Bhardwaj, 2026
Read on arXiv→