design-systems

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

• Third-party content exposure detected (high risk: 0.85). SKILL.md:164-248 describes a multi-source import flow where the agent ingests brand assets from outsider-provided inputs like Website URL, GitHub repo, documents, and Figma links (public web content / third-party authored files) via import actions, which then get synthesized into DesignSystemData and used in subsequent LLM prompts.