designcraft-catalog

Pass

Audited by Gen Agent Trust Hub on Nov 1, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external product data including SKUs, names, and specifications which may contain embedded instructions.
  • Ingestion points: Product tables, SKU lists, and files in the catalog-input directory.
  • Boundary markers: None specified in instructions for handling untrusted text data.
  • Capability inventory: Includes file creation, document modification, and file packaging via designcraft-cli (SKILL.md, references/craft-methods.md).
  • Sanitization: No sanitization or validation of the product data is performed before interpolation into the catalog.
  • [DYNAMIC_EXECUTION]: The skill generates a configuration script at runtime that is then consumed by an executable.
  • The script scripts/write_script.py compiles JSON recipes into a steps.json file.
  • This generated file is subsequently executed by designcraft-cli to perform document operations.
  • [EXTERNAL_DOWNLOADS]: The skill references an external binary download for its core functionality.
  • The file references/craft-command-evidence.json contains a link to a zip archive on GitHub (https://github.com/storytold/designcraft/releases/download/v0.8.0/designcraft-0.8.0-windows-x64-portable.zip).
  • Evidence includes a SHA256 hash for verification: 522dea416d68e5de5ec03ceede46854022587e801c76ad7a3c48b1e5e57b953a.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 1, 2026, 04:40 PM