Multi-Agent Orchestration
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely comprised of documentation and metadata. No executable scripts, binaries, or automated command execution logic are included.- [PROMPT_INJECTION]: The skill provides prompt templates for delegating tasks to subagents. These templates follow best practices for context setting and scoping, and do not contain any instructions to bypass safety filters or override system constraints.- [INDIRECT_PROMPT_INJECTION]: The skill describes patterns for synthesizing data from multiple agents, which constitutes an indirect prompt injection attack surface (Category 8). However, it proactively addresses this risk by documenting mitigation strategies, including the use of 'Critic' agents for validation and conflict resolution strategies.- [EXTERNAL_DOWNLOADS]: The skill references established orchestration frameworks such as LangGraph, AutoGen, and CrewAI for informational purposes, but does not include any automated installation or download routines.- [COMMAND_EXECUTION]: The documentation mentions the 'runSubagent' tool used in VS Code Copilot environments. This is a standard platform capability and the examples provided demonstrate safe and specific usage patterns.
Audit Metadata