learning-from-exemplar-repos

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

• Third-party content exposure detected (high risk: 0.75). The required workflow explicitly instructs using external exemplar repositories (e.g., tanstack.com, OpenPanel, opencode) and studying their route/layout/SSR/component text, which at runtime would require fetching/reading outsider-authored free-form prose/code from public repos into the agent’s LLM context (indirect prompt-injection risk).