05-ad-copy-global
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill operates solely as a text-generation assistant for marketing purposes. It does not include executable scripts, shell commands, or network-enabled tools.
- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill reads from local context files, specifically
.agents/product-marketing-context.mdand.agents/personal-brand-context.md. This is a standard pattern for personalizing agent responses based on user-provided project data and does not involve sensitive system files or network exfiltration. - [INDIRECT_PROMPT_INJECTION]: The skill has a defined attack surface for indirect prompt injection as it ingests untrusted data from product marketing context files and user-provided descriptions. However, since the skill's capabilities are limited to generating ad copy text and it lacks high-risk tools (file writing, shell execution, or network requests), the security risk is negligible.
- Ingestion points: Product marketing/personal brand context files and user responses to information-gathering questions.
- Boundary markers: Absent; the skill relies on the agent's internal instruction following.
- Capability inventory: Text generation only; no subprocess calls, file-writes, or network operations.
- Sanitization: Absent.
Audit Metadata