wear-the-hat

Pass

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill's primary function is to facilitate role-playing and task delegation. It uses a transparent, table-based heuristic (references/auto-pick-heuristic.md) to map keywords to roles defined in a separate catalog (team-composer/references/role-personas.md).
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests user-provided task briefs and incorporates them into persona-driven prompts or sub-agent instructions.
  • Ingestion points: User task briefs processed in SKILL.md (Phase 1 and Phase 4).
  • Boundary markers: The sub-agent brief template in Phase 4 uses structured sections (Task, Scope, Constraints) to delimit user input from instructions.
  • Capability inventory: The skill can execute shell commands (inline mode) or spawn sub-agents through sub-agent-coordinator.
  • Sanitization: No explicit sanitization of user-provided content is mentioned, but the skill operates within the agent's standard safety guardrails and does not request elevated privileges.
  • [EXTERNAL_DOWNLOADS]: The skill documentation describes installation via npx skills (a well-known service from Vercel Labs) and references dependencies on other skills authored by the same vendor (sorawit-w), such as team-composer and sub-agent-coordinator. These references are standard for the platform's ecosystem and do not involve untrusted sources.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 19, 2026, 07:12 AM
Security Audit — agent-trust-hub — wear-the-hat