stackblitz-sdk-patterns

Pass

Audited by Gen Agent Trust Hub on Sep 10, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill establishes architectural boundaries for managing WebContainer startup and teardown, replacing race-prone concurrent boots with a single idempotent adapter promise.- [EXTERNAL_DOWNLOADS]: The skill references and fetches technical documentation from the official webcontainers.io domain, which is the established service for the technology described.- [INDIRECT_PROMPT_INJECTION]: The skill includes an ingestion surface by analyzing local application code via Read, Glob, and Grep tools to map SDK integration points. However, it specifically mandates that commercial keys and authentication secrets must not be exposed through state or logs, and the workflow is focused on structural refactoring without unsafe data interpolation.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 10, 2026, 09:27 AM
Security Audit — agent-trust-hub — stackblitz-sdk-patterns