skills/mukul975/anthropic-cybersecurity-skills/exploiting-template-injection-vulnerabilities/Gen Agent Trust Hub
exploiting-template-injection-vulnerabilities
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill represents standard cybersecurity testing functionality. All documented exploitation payloads (such as Jinja2, Twig, and Freemarker expressions) are intended to execute on the remote target under evaluation, not on the host executing the skill. No security boundaries are bypassed locally.
Audit Metadata