solo-customer-finder

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

• Third-party content exposure detected (high risk: 0.78). The required workflow explicitly performs runtime web fetching/search of public sources (e.g., forums/Reddit/HN/GitHub/company pages) and then ingests “exact evidence quote (minimal) or paraphrase” from those outsider-authored pages into the agent’s LLM context for scoring and report writing, which is indirect prompt-injection exposure.