coordinator

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill provides high-level architectural instructions for agent orchestration and task management without including executable code, sensitive file access, or external network operations.
  • [INDIRECT_PROMPT_INJECTION]: The skill describes a workflow that processes outputs from other agent workstreams, which represents an indirect prompt injection surface.
  • Ingestion points: The skill observes returned results and artifacts from delegate assignments (SKILL.md).
  • Boundary markers: Present; the instructions require the agent to check results against the assignment rather than trusting success claims (SKILL.md).
  • Capability inventory: None; the skill provides instructions but no scripts with subprocess calls, exec/eval, or network operations are included in this skill.
  • Sanitization: Absent; the workflow relies on logical verification and QA rather than technical escaping or filtering of content.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 12:14 AM
Security Audit — agent-trust-hub — coordinator