competitive-platform-analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.85). The skill’s runtime workflow explicitly instructs gathering competitor information from outsider-authored public sources (e.g., “Portfolio / craft platforms,” “Awards / curated showcases,” “Competitor's own site,” “LinkedIn,” “Review directories,” “Open / public work”), which are then ingested as readable text into the agent’s LLM context for scoping/tiering.