peach-agent-team

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly loads and analyzes arbitrary Figma URLs using the FigmaRemote MCP (see references/figma-workflow.md and the ui-dev workflows in SKILL.md), fetching third-party/user-generated design content that the agent reads and uses to drive UI generation and subsequent actions, which could carry indirect prompt injection risk.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly loads Figma files at runtime via calls like mcp__FigmaRemote__get_figma_data(url="https://figma.com/file/xxx") / https://www.figma.com/file/[FILE_ID]/[FILE_NAME], and that fetched design data is used to direct the UI-generation prompts and behavior, so the external Figma URL can directly control agent instructions.