ai-personalization-prompts

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

• Third-party content exposure detected (medium risk: 0.65). Although the skill’s runtime path prefers ColdIQ MCP/API enrichment, it explicitly defines a “no ColdIQ access” fallback that uses a manual Claygent crawl; that workflow would fetch outsider-authored free text from third-party web sources/pages and inject it into the LLM context for prompt variable generation.