orchestrating-multi-agent-systems

Pass

Audited by Gen Agent Trust Hub on May 4, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill implements a multi-agent orchestration architecture where user-supplied data is passed between specialized agents. This design creates a surface for indirect prompt injection if untrusted data contains instructions that could manipulate the behavior of downstream agents.
  • Ingestion points: User inputs are received by the coordinator in assets/example_coordinator.ts and processed through functions in references/examples.md.
  • Boundary markers: The provided implementation templates do not include explicit delimiters or instructions to ignore potential injection content within user data.
  • Capability inventory: The skill specifies access to file system tools and Bash for package management via the allowed-tools field in SKILL.md.
  • Sanitization: No input sanitization or escaping logic is demonstrated in the provided code examples.
  • [EXTERNAL_DOWNLOADS]: The skill instructions and documentation guide users to install standard AI SDK packages from official registries to support the multi-agent system's functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
May 4, 2026, 03:15 AM
Security Audit — agent-trust-hub — orchestrating-multi-agent-systems