tzai-storybook

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill searches for an executable script named tzai-image within several standard local agent directories (e.g., ~/.agents/skills/, ~/.claude/skills/). It then executes this script using the bash interpreter to process the image generation request.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied text to generate image prompts, creating a potential surface for indirect injection.
  • Ingestion points: User input is captured and passed directly to the engine's --prompt argument.
  • Boundary markers: The script uses double quotes to wrap the user-supplied subject, providing basic shell-level delimitation.
  • Capability inventory: The skill has the capability to execute shell commands via the identified engine script.
  • Sanitization: There is no explicit sanitization or filtering of the user input beyond standard shell quoting.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 02:50 AM
Security Audit — agent-trust-hub — tzai-storybook