competitive-intelligence

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

• Third-party content exposure detected (high risk: 0.85). SKILL.md describes runtime web research and pulling competitor/customer text from public sources (e.g., product pages, pricing, reviews, release notes) which would be ingested into the agent’s LLM context for summarization and battlecard generation, creating an indirect prompt-injection risk from outsider-authored free text.