tap-explorer
Pass
Audited by Gen Agent Trust Hub on Jun 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a structured methodology for code analysis, optimization, and security auditing. It uses a Tree of Attacks with Pruning (TAP) pattern to systematically explore solution vectors. No malicious instructions, unauthorized file access, or data exfiltration patterns were detected.
- [PROMPT_INJECTION]: The skill involves generating prompts based on user goals and source code context. While this creates a theoretical surface for indirect prompt injection (Category 8), it is the primary intended function of the tool for code analysis. There are no instructions to bypass safety filters or ignore system constraints.
- [EXTERNAL_DOWNLOADS]: The skill includes an attribution link to the ZeroLeaks TAP methodology repository. This is a documentation reference and does not involve runtime code execution or untrusted downloads.
Audit Metadata