brainstorm-feature

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

• Third-party content exposure detected (high risk: 0.95). The workflow explicitly performs runtime web research by calling google_web_search and then read_url_content to ingest arbitrary third-party engineering/compliance articles into the agent’s LLM context (public web content fetched at runtime).

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs runtime fetching via read_url_content/google_web_search to ingest external web pages (e.g., AWS Architecture Center https://aws.amazon.com/architecture or engineering blogs like https://netflixtechblog.com) and then synthesize design guidance from that content, meaning remote pages directly influence agent prompts and behaviour.