frappe-ops-bench

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The prompt includes many example commands and config entries containing plaintext passwords and keys (e.g., --admin-password admin, --mariadb-root-password rootpass, --backup-encryption-key mykey and db_password/FRAPPE_DB_PASSWORD), which would require the LLM to reproduce secret values verbatim in generated commands or code.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill includes a runtime command that fetches code from an external Git URL (e.g., "bench get-app https://github.com/org/custom-app.git --branch main"), which at runtime clones remote repository content that can contain arbitrary Python/CLI hooks executed during install or when bench discovers and runs app-provided commands, so the fetched content can directly execute code and is a required dependency for installing/using that app.

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

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The prompt contains many commands that require or suggest using sudo and modifying system services and configuration (bench setup production, setup nginx, reload nginx, restart services), as well as destructive actions like drop-site and creating sites/users — all of which change machine state and need elevated privileges.