anima-core-workflow-b

Pass

Audited by Gen Agent Trust Hub on Apr 13, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill utilizes the official @animaapp/anima-sdk and references documentation from docs.animaapp.com. These are legitimate resources associated with the Anima design platform.\n- [COMMAND_EXECUTION]: Employs standard Node.js fs module functions, such as mkdirSync and writeFileSync, to organize and save generated React/HTML components locally.\n- [SAFE]: Handles sensitive authentication tokens securely via process.env.ANIMA_TOKEN, avoiding hardcoded secrets in the source code.\n- [SAFE]: The skill processes data from cloned websites to generate and transform code, which constitutes an indirect prompt injection surface. However, the logic uses deterministic regex transformations and local file operations rather than passing untrusted content directly to LLM instructions.\n
  • Ingestion points: File content arrays in customizeGeneratedCode and organizeGeneratedFiles.\n
  • Boundary markers: Not present in the provided utility functions.\n
  • Capability inventory: Local file writing (fs.writeFileSync) and directory creation (fs.mkdirSync).\n
  • Sanitization: No explicit sanitization of website content is performed beyond standard string replacements.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 13, 2026, 09:30 PM