ponytail-review
Pass
Audited by Gen Agent Trust Hub on Aug 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security risks were detected in the skill instructions.
- [PROMPT_INJECTION]: The skill includes instructions to revert to 'normal mode' or 'stop ponytail-review', which are standard state-management instructions for task-specific AI agents and do not target system-level safety protocols.
- [DATA_EXFILTRATION]: The skill does not utilize network tools (such as curl or wget) or instructions to transmit data to external servers.
- [COMMAND_EXECUTION]: The skill is designed for static analysis of provided text (diffs) and does not contain commands or scripts that would lead to arbitrary code or shell execution.
- [INDIRECT_PROMPT_INJECTION]: While the skill processes untrusted user data (code diffs), it lacks any tools or capabilities (e.g., shell access, file writes, network requests) that could be exploited via embedded instructions in the processed data.
Audit Metadata