skills/itamaker/skills/stitch/Gen Agent Trust Hub

stitch

Pass

Audited by Gen Agent Trust Hub on Sep 29, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads the @google/stitch-sdk package from the public npm registry. The source organization is categorized as trusted.
  • [COMMAND_EXECUTION]: The scripts/stitch.mjs runner script executes the npm install command to bootstrap the necessary design SDK into a local cache directory.
  • [DYNAMIC_EXECUTION]: The utility uses dynamic Node.js import() calls to load the Stitch SDK from a path resolved at runtime in the user's cache folder.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides an interface to pass user-defined text prompts to the Stitch service for design generation, which is a standard surface for indirect injection.
  • Ingestion points: Input is accepted through the --prompt and --prompt-file arguments in the CLI runner.
  • Boundary markers: The skill does not implement specific delimiters or instructions to isolate the user input from the service's internal prompt logic.
  • Capability inventory: The skill can perform network requests (downloading design assets), write files (config and design output), and execute subprocesses (npm).
  • Sanitization: The prompts are passed to the SDK without explicit sanitization or validation filtering.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 29, 2026, 10:09 AM
Security Audit — agent-trust-hub — stitch