designcraft-long-document

Pass

Audited by Gen Agent Trust Hub on Nov 1, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references a specific, versioned binary of the DesignCraft CLI (v0.8.0) hosted on GitHub Releases. This reference in references/craft-command-evidence.json includes a SHA256 checksum for integrity verification, following security best practices.
  • [COMMAND_EXECUTION]: The skill utilizes a local Python utility (scripts/write_script.py) to compile layout recipes into structured command sequences. It also instructs the user on how to install related skills using the npx skills add platform command.
  • [DYNAMIC_EXECUTION]: The provided Python script performs runtime resolution of cross-references within JSON templates to generate the final execution plan for the DesignCraft CLI. This is a localized, deterministic transformation of static data.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes document content and headings. It mitigates the risk of instructions embedded in the document text by enforcing a style-based selection logic (targeting paragraph styles rather than raw text matching) and requiring the agent to perform structural verification of the final output.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 1, 2026, 04:43 PM