designcraft-newsletter

Warn

Audited by Gen Agent Trust Hub on Nov 1, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The file references/craft-command-evidence.json specifies a remote download for a portable Windows binary from https://github.com/storytold/designcraft/releases/download/v0.8.0/designcraft-0.8.0-windows-x64-portable.zip. This repository is not associated with a known trusted vendor.
  • [COMMAND_EXECUTION]: The skill uses designcraft-cli.exe to perform document layout and export operations as described in references/craft-methods.md.
  • [DYNAMIC_EXECUTION]: The script scripts/write_script.py generates a JSON-based command script (steps.json) at runtime by resolving template references. This generated file is subsequently passed to the designcraft-cli for execution.
  • [INDIRECT_PROMPT_INJECTION]:
  • Ingestion points: The skill ingests raw user manuscript text via the text parameter in frame.create calls (e.g., in references/craft-steps.json).
  • Boundary markers: No delimiters or safety instructions are used to distinguish user-provided content from agent instructions during the layout process.
  • Capability inventory: The skill has the ability to create, modify, and save files to the local file system using designcraft-cli, including PDF and IDML exports.
  • Sanitization: There is no evidence of sanitization or character escaping for the user-provided text before it is inserted into the document frames.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Nov 1, 2026, 04:40 PM