vue-architect

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

• Third-party content exposure detected (medium risk: 0.65). The skill reads outsider-authored free text into the agent context because SKILL.md includes multiple runtime-possible external “Source” URLs (e.g., Medium article, EU policy page) and the workflow implies ingesting these sources for guidance via retrieval/RAG, which would fetch public web content authored by third parties.