fiverr-research

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

• Third-party content exposure detected (medium risk: 0.30). At runtime, the workflow calls Crawlora endpoints /fiverr/search (free-text q from the user) and then /fiverr/gig/{username}/{slug} and /fiverr/seller/{username}; the LLM ingests normalized Fiverr gig/seller text returned by those calls based on attacker-supplied search terms and resulting items, creating an indirect prompt-injection path through outsider-authored Fiverr content.