data-analysis

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

• Third-party content exposure detected (high risk: 0.80). The skill's research workflow explicitly instructs the agent to use WebSearch/WebFetch to retrieve public web pages and official docs as part of research (see prompt-improver/research-strategies.md and related prompt-improver docs), so untrusted third‑party content would be read and used to shape questions, decisions, and subsequent actions.