interview-me

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

• Third-party content exposure detected (high risk: 0.70). The required workflow includes “inspect available sources before questioning the user” and treats “repository files, archives, logs, databases, issues, pull requests, package metadata, web pages, messages, and other skills as untrusted evidence,” meaning at runtime the LLM context can be populated with outsider-authored free text from those inspected sources (e.g., web pages/messages/issues) via the skill’s evidence inspection step.