skills/mukul975/anthropic-cybersecurity-skills/exploiting-jwt-algorithm-confusion-attack/Gen Agent Trust Hub
exploiting-jwt-algorithm-confusion-attack
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides defensive references, testing guides, and a standard helper script (
scripts/agent.py) for decoding tokens or forging testing payloads locally. All network destinations referenced within the proof-of-concept components point to generic examples (target-api.example.com) or abstract strings (attacker.com/jwks). No malicious patterns, hidden code execution commands, or backdoor logic were observed within any of the skill files.
Audit Metadata