zmx
Pass
Audited by Gen Agent Trust Hub on Aug 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns or security vulnerabilities were identified. The skill facilitates standard development workflows with appropriate project-level isolation using git-based session naming conventions.\n- [COMMAND_EXECUTION]: The skill provides patterns for starting and stopping background processes. This functionality is scoped to the development environment and used for its primary purpose of managing servers and watchers.\n- [PROMPT_INJECTION]: The skill defines patterns for reading process output, which establishes a surface for indirect prompt injection.\n
- Ingestion points: Process logs read via
zmx history(SKILL.md).\n - Boundary markers: Absent in the demonstrated shell commands.\n
- Capability inventory: Command execution capability through
zmx runand shell access.\n - Sanitization: Absent; the examples rely on simple pattern matching (e.g., using grep or ripgrep) to verify process state.
Audit Metadata