product-marketing-context

Pass

Audited by Gen Agent Trust Hub on Apr 16, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect Prompt Injection Surface. The 'auto-draft' feature reads repository files like README, package.json, and marketing copy to automatically generate context. This creates a surface where instructions embedded in these files could influence the agent's behavior during the drafting process. However, the risk is minimal as the skill's capabilities are limited to writing a markdown file. * Ingestion points: Codebase files including README and package.json (SKILL.md). * Boundary markers: None identified for isolating ingested content. * Capability inventory: File write to .agents/product-marketing-context.md. * Sanitization: None performed on ingested content.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 16, 2026, 07:42 PM
Security Audit — agent-trust-hub — product-marketing-context