NYC

python-expert

Fail

Audited by Gen Agent Trust Hub on Feb 15, 2026

Risk Level: HIGHPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • PROMPT_INJECTION (HIGH): The skill is susceptible to Indirect Prompt Injection (Category 8) as it is designed to analyze external, untrusted Python code.\n
  • Ingestion points: Python source code, docstrings, and comments provided by users for review, debugging, or refactoring (SKILL.md).\n
  • Boundary markers: Absent. The instructions do not define delimiters or provide warnings to the agent to disregard instructions within the code being analyzed.\n
  • Capability inventory: The skill allows access to Write and Bash(python:*) tools, enabling filesystem modification and arbitrary code execution.\n
  • Sanitization: Absent. There is no mechanism to sanitize or validate the external content before it is processed by the LLM.\n- COMMAND_EXECUTION (MEDIUM): Allowing the Bash(python:*) tool provides a direct vector for code execution. While intended for legitimate debugging, it can be abused if the agent's logic is manipulated through instructions injected via code comments or strings.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 15, 2026, 09:41 PM