tiktok-account-search

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

• Third-party content exposure detected (high risk: 0.75). Runtime path: the script calls a public third-party API endpoint (https://redfox.hk/.../searchUser) with the user keyword, then the LLM ingests the returned JSON fields like nickname and signature (free-form bios) from that outsider-controlled source into the agent context.