copywriting
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill implements robust defensive measures by explicitly instructing the agent to treat all external inputs—such as scraped web pages, reviews, comments, and emails—as untrusted and to specifically disregard any instructions found within that content.
- [DATA_EXFILTRATION]: While the skill defines integration layers for various marketing and CRM platforms (e.g., HubSpot, Salesforce, Google Ads), it mandates that persistent writes or account mutations require explicit user authorization and emphasizes keeping credentials out of the skill source code.
- [REMOTE_CODE_EXECUTION]: No patterns involving remote code execution, piped shell commands, or unverified package installations were detected. The mentioned connectors refer to well-known services and standard integration layers.
- [COMMAND_EXECUTION]: There are no instances of direct shell command execution or privilege escalation attempts. The skill operates within the scope of document processing and artifact generation.
Audit Metadata