design-system

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

• Third-party content exposure detected (high risk: 0.75). The required workflow reads multiple outsider-authored markdown files from the workspace at runtime (e.g., design/gdd/systems-index.md, dependency GDDs, and docs/consistency-failures.md) and injects their extracted prose into the agent’s LLM context during “Phase 2 (Gather Context)”.