designing-in-figma

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

• Third-party content exposure detected (medium risk: 0.30). In the designing-in-figma workflow, the agent ingests outsider-authored free text at runtime when it reads a user-selected Claude Design project via DesignSync (list_projects→list_files→get_file) and then parses files like _ds_manifest.json and component .prompt.md/.html as data for generating Figma content.

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

• Potentially malicious external URL detected (high risk: 0.90). This skill explicitly instructs the agent to "load" required sub-skills from the remote GitHub path https://github.com/figma/mcp-server-guide/tree/main/skills at runtime and to delegate to them, meaning external content at that URL will be fetched and can directly control agent instructions during execution.