maestro-tools-execute

Pass

Audited by Gen Agent Trust Hub on Jul 3, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill is designed to execute shell commands, file operations, and code changes as defined in external tool specification documents. This behavior is mediated through the maestro command-line tool.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes instructions from external 'knowhow' and 'spec' documents. While this creates a surface for indirect prompt injection, the skill mitigates this risk through 'Phase Gate 3', which requires explicit user confirmation via request_user_input before execution begins, and provides a 'View only' mode for safe inspection of steps.
  • Ingestion points: Tool specifications are loaded via maestro search and maestro load based on user-provided arguments (SKILL.md).
  • Boundary markers: Includes an 'Output boundary' notice and sequential execution invariants.
  • Capability inventory: Uses Bash, Write, and Edit tools to perform actions defined in the loaded specs.
  • Sanitization: Relies on human-in-the-loop review and explicit confirmation before execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 3, 2026, 08:44 AM
Security Audit — agent-trust-hub — maestro-tools-execute