start-leader
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions strictly focus on orchestration and coordination within an agent swarm using established MCP tools.
- [INDIRECT_PROMPT_INJECTION]: As a coordinator agent, this skill inherently creates a surface for indirect prompt injection by ingesting and acting upon data from multiple agents and inbox messages.
- Ingestion points: File SKILL.md specifies the agent should monitor messages via
read-messagesand check task progress viaget-task-details. - Boundary markers: The skill does not define specific delimiters for separating untrusted data from instructions.
- Capability inventory: The agent has the ability to assign tasks (
send-task), manage the task pool (task-action), and communicate through swarm-wide chat. - Sanitization: No explicit validation or sanitization routines for external or worker-provided content are described.
Audit Metadata