NYC

skill-creator-plus

Fail

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: HIGHCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [Indirect Prompt Injection] (HIGH): The skill's primary function includes reviewing external, untrusted SKILL.md files, which creates a significant attack surface for indirect prompt injection.\n
  • Ingestion points: Specifically identifies 'Reviewing existing SKILL.md files' and references 'references/skill-review-checklist.md' as targets for processing.\n
  • Boundary markers: Absent. The instructions do not provide delimiters or warnings to the agent to treat the reviewed content as untrusted data.\n
  • Capability inventory: The skill explicitly allows for the execution of Python scripts (init_skill.py and package_skill.py) which could be targeted by an injection attack.\n
  • Sanitization: Absent. There is no evidence of input validation or content filtering for the files being analyzed.\n- [Command Execution] (MEDIUM): The creation workflow instructs the agent to run local scripts (scripts/init_skill.py, scripts/package_skill.py) using user-defined variables such as <name> and <path>. This pattern is susceptible to command injection if the agent does not strictly validate these inputs before invoking the subprocess.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 16, 2026, 11:34 AM