frappe-ops-app-lifecycle

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes an example that embeds a GitHub token directly in a URL (bench get-app https://TOKEN@github.com/...), which instructs including secrets verbatim in commands/requests and is therefore high-risk credential handling.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 1.00). The skill's required workflows (e.g., "3. Getting Apps from Git" and "Workflow 2: Install an Existing App from GitHub") instruct using bench get-app to clone public GitHub repositories whose untrusted, user-provided code (hooks, patches, migrations) is then installed/executed (bench install-app / migrate), so third-party content is fetched and can materially influence runtime actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly runs runtime fetches of git repositories (e.g., "bench get-app https://github.com/org/my_app" and "bench get-app git@github.com:org/private_app.git"), and the cloned repositories can contain hooks, patches and after_install code that is executed during install/migrate, so remote content can execute code at runtime.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The skill includes explicit sudo commands (e.g., sudo bench setup production USERNAME, sudo supervisorctl restart all, sudo systemctl restart supervisor) and guidance to configure system services (nginx, supervisor) that modify system-level configuration and require elevated privileges, so it can change the machine's state and should be flagged.