product-marketing-context

Pass

Audited by Gen Agent Trust Hub on Aug 11, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill's primary function is to create and update a markdown file at .agents/product-marketing-context.md. This is a standard practice for maintaining local agent state.
  • [COMMAND_EXECUTION]: The skill uses automated codebase analysis (scanning README, package.json, and marketing files) to draft content. This is a common and safe capability for development-focused agents.
  • [DATA_EXPOSURE_&_EXFILTRATION]: While the skill reads project files (README, landing pages, package.json) to generate marketing copy, it does not attempt to access sensitive system files, credentials, or perform unauthorized network operations.
  • [PROMPT_INJECTION]: No patterns of instruction overrides, safety filter bypasses, or adversarial roleplay were detected in the instructions.
  • [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect injection if the files it reads (like a README or landing page) contain malicious instructions. However, the risk is low as the output is primarily structured markdown text for human review.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 11, 2026, 09:01 PM
Security Audit — agent-trust-hub — product-marketing-context