@tank/figma-to-code
Pass
Audited by Gen Agent Trust Hub on May 31, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill processes external data from Figma via MCP tools (SKILL.md, references/figma-extraction-workflow.md). While this provides a surface for indirect prompt injection, the risk is mitigated as the skill manifest (tank.json) denies high-privilege permissions like file-write, outbound network, and subprocess execution. Mandatory Evidence Chain: 1. Ingestion points: Figma node metadata and properties; 2. Boundary markers: Absent; 3. Capability inventory: Broad read-only access ('**/*') with no write, network, or subprocess permissions; 4. Sanitization: Not specified for external design content.
- [EXTERNAL_DOWNLOADS]: The documentation mentions using Google Fonts and Figma assets. These references target trusted organizations (Google) and well-known services (Figma) and do not pose a security risk.
- [SAFE]: The skill requests broad read access to the filesystem. This is used to analyze existing project components and design tokens to ensure implementation consistency, which is associated with the primary purpose of the skill.
Audit Metadata