clifton-sellers-content-prompts
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE] (SAFE): The skill consists entirely of markdown files (
SKILL.mdand reference files). No executable code (Python, Node.js, Shell), configuration files, or build scripts are present. - [DATA_EXFILTRATION] (SAFE): There are no network request commands (
curl,wget,fetch) or sensitive file path references. The skill does not have the capability to move data externally. - [PROMPT_INJECTION] (SAFE): The prompts utilize role-play personas (e.g., 'adversarial business strategist', 'precision editor'). These instructions are confined to the intended business analysis use case and do not contain patterns aimed at bypassing LLM safety filters or extracting system prompts.
- [COMMAND_EXECUTION] (SAFE): No shell command execution or subprocess spawning mechanisms are identified.
- [REMOTE_CODE_EXECUTION] (SAFE): There are no patterns involving the download or execution of remote scripts.
Audit Metadata