Skills
Navigate
skills/zseven-w/craft-skills/single-path-process-diorama/Gen Agent Trust Hub

single-path-process-diorama

Pass

Audited by Gen Agent Trust Hub on Sep 23, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests user-supplied text describing 3–5 step processes to generate image creation prompts, creating an indirect prompt injection surface.
  • Ingestion points: User-provided process descriptions processed according to instructions in SKILL.md.
  • Boundary markers: SKILL.md mandates the creation of a 'scene-contract.md' to define and lock process steps before image generation.
  • Capability inventory: SKILL.md describes image generation tool calls, and README.md describes local Python script execution.
  • Sanitization: SKILL.md provides strict mapping rules for user steps to visual macro-stages to prevent arbitrary modification.
  • [COMMAND_EXECUTION]: Instructions in README.md and SKILL.md prompt the execution of the included scripts/verify_delivery.py script to validate PNG file structure and ensure mandatory documentation is present.
  • [DYNAMIC_EXECUTION]: The regression test suite in tests/test_verify_delivery.py utilizes the importlib.util module to dynamically load and execute the local verification script for automated testing.
  • [EXTERNAL_DOWNLOADS]: The README.md file provides instructions to clone the skill source code from the author's GitHub repository at github.com/ZSeven-W/craft-skills.git during the installation process.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 23, 2026, 06:21 PM