stitch-code-to-design

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill references source material from the Google organization's GitHub repository for maintenance and version synchronization purposes.
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze untrusted project files including configuration files and component templates.
  • Ingestion points: Data enters the agent's context through local file reading of the application's source code and build metadata.
  • Boundary markers: The instructions lack specific boundary markers or warnings to disregard potential embedded instructions within the project source code.
  • Capability inventory: The workflow includes file creation and network communication for uploading design assets to the Stitch platform.
  • Sanitization: No explicit requirements for sanitizing or validating ingested code snippets are specified before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 03:52 AM
Security Audit — agent-trust-hub — stitch-code-to-design