content-research
Pass
Audited by Gen Agent Trust Hub on Feb 19, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE] (SAFE): The skill consists entirely of markdown files (.md). There are no Python, JavaScript, or Shell scripts, which eliminates the risk of local code execution, system modification, or persistence.
- [EXTERNAL_DOWNLOADS] (SAFE): No remote URLs are used for script downloads or dynamic execution. Referenced domains (e.g., Perplexity, ChatGPT) are discussed conceptually as targets for optimization, not as sources of external code.
- [CREDENTIALS_UNSAFE] (SAFE): No hardcoded secrets, API keys, or private tokens were identified within the templates or guides.
- [PROMPT_INJECTION] (SAFE): The documentation focuses on influencing external AI models through legitimate content formatting (SEO-style). There are no instructions targeting the agent's internal logic, safety constraints, or attempting to extract system prompts.
Audit Metadata