ponytail

Pass

Audited by Gen Agent Trust Hub on Jun 28, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill consists exclusively of instructional text in markdown and a diff patch file. It does not contain any executable scripts, binary files, or external dependencies.
  • [PROMPT_INJECTION]: The skill uses persistence instructions ('ACTIVE EVERY RESPONSE') and deactivation triggers ('stop ponytail') to manage its persona. These are standard behavioral framing techniques for specialized agent roles and are accompanied by explicit mandates to never simplify away security measures or input validation.
  • [NO_CODE]: No source code, package manifests, or remote code patterns were detected. The skill functions solely as a logical constraint for the AI agent's response patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 28, 2026, 05:51 PM
Security Audit — agent-trust-hub — ponytail