consultancy-practices
Pass
Audited by Gen Agent Trust Hub on Mar 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill "consultancy-practices" consists of high-level professional guidelines and markdown-based instructions for agent behavior. No executable scripts or binary files are included.
- [SAFE]: No hardcoded credentials, sensitive file access, or unauthorized network operations were identified in the provided files.
- [SAFE]: The 'Agent Integration' and 'Fleet Coordination' sections provide documentation for orchestrating tasks and managing agent fleets, which align with the skill's primary purpose of professional quality assessment.
- [SAFE]: An analysis of indirect prompt injection surfaces (Category 8) shows that while the skill processes external client data, this is necessary for its function as a consultancy tool.
- Ingestion points: Processes interview notes, deployment walkthroughs, and bug reports (SKILL.md).
- Boundary markers: None explicitly defined for untrusted input.
- Capability inventory: Orchestrates codebase assessments and ROI calculations via Task() and FleetManager calls (SKILL.md).
- Sanitization: No sanitization is present, but given the professional advisory nature and lack of malicious logic, this remains a low-risk surface.
Audit Metadata