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

stitch

Pass

Audited by Gen Agent Trust Hub on Nov 6, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONCREDENTIALS_UNSAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads the official @google/stitch-sdk package from the npm registry. This package is maintained by a trusted organization and is used as the primary engine for design generation and management.
  • [COMMAND_EXECUTION]: The skill's runner executes npm install to handle the installation and updates of the SDK. It mitigates potential risks by using the --ignore-scripts flag, ensuring that no arbitrary post-install scripts from the package or its dependencies are executed.
  • [DYNAMIC_EXECUTION]: After downloading the SDK to a local cache, the script dynamically loads the library using the import() function. This is a standard and safe approach for handling runtime-installed dependencies in a Node.js environment.
  • [CREDENTIALS_UNSAFE]: The skill manages authentication using API keys and OAuth tokens. It implements necessary security controls, such as a redactSecrets function that masks sensitive values in console output, and provides clear documentation on managing secrets via environment variables or git-ignored configuration files.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests user prompts and external design metadata. It uses structured data formats (JSON) and clear boundary markers in its internal logic to process this information, relying on the underlying platform's safety guidelines to handle natural language instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 6, 2026, 04:07 AM