skills/sanxzy/skills/qa-test/Gen Agent Trust Hub

qa-test

Pass

Audited by Gen Agent Trust Hub on May 13, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The scripts/preflight.sh script installs the chrome-devtools-axi and agent-browser CLI tools globally using npm if they are not detected on the system. These tools are the primary engines for the skill's QA operations.
  • [COMMAND_EXECUTION]: The skill orchestrates complex browser interactions by executing shell commands and scripts through the two browser drivers. This includes navigating pages, filling forms, and running performance audits.
  • [DATA_EXFILTRATION]: The skill enforces a security boundary by requiring all test artifacts (screenshots, network logs, reports) to be stored in a local directory rather than being outputted directly to the conversation, mitigating the risk of unintentional data exposure.
  • [CREDENTIALS_UNSAFE]: The skill provides explicit instructions and tools (auth vault) to handle credentials securely, specifically warning against inlining secrets in executable scripts or command arguments.
Audit Metadata
Risk Level
SAFE
Analyzed
May 13, 2026, 07:36 AM
Security Audit — agent-trust-hub — qa-test