maestro-core

Pass

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: LOW
Full Analysis
  • [Indirect Prompt Injection] (LOW): This skill acts as a router that maps user input (triggers like ds, ci, fb) to agent actions (loading sub-skills via the skill() tool). This is an inherent part of its orchestration purpose.
  • Ingestion points: User-supplied trigger phrases and slash commands defined in the 'Quick Triggers' and 'Ownership Matrix' tables.
  • Boundary markers: Absent; the skill relies on direct matching of user triggers.
  • Capability inventory: Instructs the agent to invoke the skill() tool to load relative local skill paths (e.g., ../designing/SKILL.md).
  • Sanitization: None specified, but the triggers are mapped to a static internal routing table.
  • [Prompt Injection] (SAFE): The instructions provide guidance on workflow state and hierarchy. There are no attempts to bypass safety constraints, extract system prompts, or override agent personality.
  • [Data Exposure] (SAFE): No hardcoded credentials (API keys, tokens) or sensitive file paths (e.g., .ssh, .env) were found in the skill definition.
  • [Remote Code Execution] (SAFE): No remote downloads (curl/wget), piped shell execution, or package installations are present. All referenced sub-skills are relative local paths.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 16, 2026, 03:19 AM