special-creative-spark

Pass

Audited by Gen Agent Trust Hub on Mar 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or security vulnerabilities were identified in either version of the skill. The content consists entirely of instructional markdown and configuration metadata.
  • [PROMPT_INJECTION]: The skill utilizes a high-pressure persona ('PUA-style') to encourage creative performance. This is a stylistic role-play choice for brainstorming and does not include instructions to bypass safety guardrails, override system prompts, or disregard safety protocols.
  • [DATA_EXFILTRATION]: The skill does not contain any network-enabled capabilities, sensitive file path references (such as SSH keys or environment variables), or credential exposure patterns.
  • [REMOTE_CODE_EXECUTION]: There are no indicators of remote code execution, external package installations, or scripts that fetch content from the web.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 14, 2026, 09:54 AM