prompt-writing

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

• Third-party content exposure detected (high risk: 0.90). The skill's research and templates (e.g., templates/research-task.md and its filled example) explicitly instruct agents to consult external public sources such as "AWS S3 Vectors documentation", "Pinecone pricing and regions page", and "Hacker News discussions", which are untrusted/third‑party content the agent is expected to read and use to make decisions (e.g., choosing a vector DB), so it can enable indirect prompt injection.