selenium

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt embeds plaintext credentials (e.g., "securePassword123") directly in example test code and demonstrates sending them verbatim (send_keys), which instructs the agent to output or reproduce secrets in cleartext.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.70). The skill clearly pulls and runs external runtime artifacts that execute remote code—for example the Docker images referenced in docker-compose (selenium/hub:4.18, selenium/node-chrome:4.18, selenium/node-firefox:4.18) and the CI service image (selenium/standalone-chrome:4.18), and it invokes external GitHub Actions (actions/checkout@v4, actions/setup-python@v5) which are fetched and executed at runtime.