skills/manaflow-ai/cmux/cmux-settings/Gen Agent Trust Hub

cmux-settings

Pass

Audited by Gen Agent Trust Hub on Oct 13, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes the cmux command-line tool for semantic validation of configuration changes and supports opening settings in user-defined editors such as VS Code, Cursor, or system defaults using environment variables like $EDITOR.
  • [DATA_EXPOSURE]: Accesses and modifies the application's configuration file located at ~/.config/cmux/cmux.json. This access is restricted to the skill's primary purpose and employs secure file handling, including writing undo receipts with restricted permissions (0600).
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses an ingestion surface through the reading of the cmux.json configuration file. It mitigates potential risks by enforcing strict JSON structure parsing and utilizing an external validator (cmux config validate) to ensure that only schema-compliant data is processed.
  • [DYNAMIC_EXECUTION]: Uses ctypes to interface with standard system libraries (libc) for performing atomic file swaps using renamex_np (macOS) and renameat2 (Linux). This is a standard implementation for ensuring file integrity during writes and does not involve untrusted code injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 13, 2026, 10:56 AM