figma-design-md-plugin
Warn
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONREMOTE_CODE_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions to download source code from an external repository at
https://github.com/bergside/design-md-figma.gitusinggit clone. - [REMOTE_CODE_EXECUTION]: The installation process involves executing
npm installandnpm run buildon the cloned repository, which runs potentially unverified code and scripts on the local machine. - [INDIRECT_PROMPT_INJECTION]: The plugin logic extracts text from Figma objects (such as style and variable names) and interpolates them directly into markdown instructions for other AI agents (
SKILL.md). If a Figma file contains malicious style names, the generated instructions could manipulate the behavior of downstream agents. - Ingestion points: Figma file metadata (
figma.root.name), page names, style names, and variable names extracted insrc/code.ts. - Boundary markers: Absent. The content is interpolated directly into the markdown output without delimiters or specific "ignore instructions" warnings.
- Capability inventory: The skill is designed to generate a
SKILL.mdfile that provides strict implementation rules for AI coding agents. - Sanitization: The code performs no sanitization or validation of the strings extracted from Figma before writing them to the generated output.
Audit Metadata