skills/testland/qa/workbox-tests/Gen Agent Trust Hub

workbox-tests

Pass

Audited by Gen Agent Trust Hub on Oct 2, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: No security issues detected. The skill provides legitimate testing patterns for PWA service workers.
  • [EXTERNAL_DOWNLOADS]: The skill references standard dependencies like @playwright/test and vitest from public registries. It also links to official Google Chrome developer documentation and GitHub repositories, which are recognized as trusted sources.
  • [COMMAND_EXECUTION]: Shell commands provided are limited to standard development and CI tasks, such as installing packages, building the project, and executing tests.
  • [INDIRECT_PROMPT_INJECTION]: The skill has a data ingestion surface where it reads and validates the manifest of a locally built service worker file.
  • Ingestion points: Reads ./dist/sw.js in SKILL.md to verify the precache manifest.
  • Boundary markers: None identified.
  • Capability inventory: Uses readFileSync, JSON.parse, and Playwright browser evaluation.
  • Sanitization: None identified, as the data is a local build artifact.
  • Analysis: This is a standard practice for verifying build-time injections and does not present an identified risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 2, 2026, 10:16 AM