skills/adobe/skills/qa/Gen Agent Trust Hub

qa

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads the axe-core library from the jsdelivr CDN to perform accessibility audits. This is a well-known and trusted service for delivering open-source libraries.
  • [COMMAND_EXECUTION]: Utilizes Playwright for automated browser testing. This is used to validate page rendering, capture screenshots for visual regression, and measure performance metrics.
  • [DATA_EXFILTRATION]: No evidence of unauthorized data access or exfiltration. The skill reads from the target site's URL and writes reports locally to the stardust/qa/ directory as specified in its documentation.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests content from external URLs to perform QA checks. It implements text normalization and tag stripping before including content in reports, which helps mitigate risks associated with processing untrusted data.
  • Ingestion points: Site content is fetched in routing.mjs, content.mjs, links.mjs, metadata.mjs, and perf.mjs via network requests and browser navigation.
  • Boundary markers: Findings are encapsulated in structured JSON (report.json) and HTML (report.html) report formats.
  • Capability inventory: The skill performs network read operations, file system writes to a specific directory, and browser automation via Playwright.
  • Sanitization: Employs stripTags and normText functions to clean and normalize text extracted from the site before reporting.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 06:01 PM
Security Audit — agent-trust-hub — qa