growth-conversion-optimization
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill contains a rule prohibiting 'adversarial prompt injection' and 'ignore previous instructions' patterns in generated marketing content. A static detector flagged these keywords; however, the analysis confirms they are used as part of a defensive guideline to prevent the AI from generating manipulative or deceptive content.
- [COMMAND_EXECUTION]: The verification section provides instructions for running local scripts (e.g.,
scripts/marketing-optimizer.py) to audit funnels and bot-readiness scores. These are standard development tools for the intended domain and do not involve remote code execution or privilege escalation. - [EXTERNAL_DOWNLOADS]: The skill references the use of OpenRouter's API for copy optimization and variant synthesis. OpenRouter is a well-known service for aggregate LLM access, and its inclusion is functional and documented neutrally.
- [DATA_EXFILTRATION]: The instructions include a specific 'Do Not' rule against logging private user prompts, file paths, raw IP addresses, or unredacted query strings in conversion telemetry, which demonstrates a commitment to data privacy and security best practices.
Audit Metadata