APEX Methodology

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

• Third-party content exposure detected (high risk: 0.80). The skill's required workflow explicitly runs a "research-expert" phase (see SKILL.md and references/01-analyze-code.md) and the design phase (references/design/02-search-inspiration.md) that instruct the agent to consult external/public sources (framework docs, 21st.dev, shadcn/ui and similar registries) and to use those findings to drive planning and code actions, so untrusted third‑party content could materially influence tool decisions and enable indirect prompt injection.