figma-to-gutenberg

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads image assets from Figma's official API endpoints (GET /v1/files/:key/images) to sideload them into the WordPress media library.
  • [DATA_EXFILTRATION]: The skill transmits usage telemetry, including site context and version information, to the developer's domain at https://www.respira.press/api/skills/track-usage.
  • [PROMPT_INJECTION]: The skill processes untrusted external data from the Figma API node tree, creating a vulnerability surface for indirect prompt injection.
    1. Ingestion points: Figma node tree properties including layer names, text characters, and styles retrieved via the REST API (SKILL.md).
    1. Boundary markers: Absent. The skill does not define explicit delimiters or instructions to ignore embedded instructions within Figma data.
    1. Capability inventory: The skill can write new block content to WordPress, sideload images, and modify Global Styles via respira_build_page, respira_sideload_image, and respira_create_design_token (SKILL.md).
    1. Sanitization: Absent. The instructions do not describe validation or sanitization of the Figma metadata before it is mapped to the block editor.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 05:51 PM
Security Audit — agent-trust-hub — figma-to-gutenberg