agent-mesh-coordinator

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill's lifecycle hooks (pre and post) execute shell commands to interact with platform-specific MCP tools for swarm initialization (swarm_init), peer messaging (daa_communication), and persistent state storage (memory_usage).
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes external task data and coordinates multiple agents through peer-to-peer communication channels without explicit input sanitization.
  • Ingestion points: The $TASK environment variable in shell hooks and peer messages received via mcp__claude-flow__daa_communication.
  • Boundary markers: Absent; the $TASK input is interpolated directly into shell command strings without delimiters.
  • Capability inventory: Includes swarm initialization, broadcast communication to all agents, proposal of consensus decisions, and persistent memory storage.
  • Sanitization: No sanitization, escaping, or validation logic is present for external task content or peer-to-peer messages.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 04:32 PM