inter-agent-comms-reviewer

What problem does it solve?

This Skill reviews the security of live agent-to-agent and agent-to-MCP-server messaging so teams can detect spoofing, tampering, replay attacks, and unsafe trust in peer messages before they cause harmful actions.

Core Features & Use Cases

• Per-edge authentication review: Checks whether each sender and receiver in an agent mesh or MCP connection properly authenticates the other side.
• Integrity and replay defense review: Verifies end-to-end signing, nonce or timestamp controls, expiry, and idempotency for side-effecting commands.
• Topology and content-handling review: Confirms only approved communication paths exist and that authenticated messages are treated as untrusted data, not instructions.
• Use Case: Review a planner-worker queue, a multi-agent orchestrator, or an agent connected to remote MCP servers to find message-layer security gaps and rank them by attack path and impact.

Quick Start

Ask the skill to review your agent communication flow for spoofing, tampering, replay, confidentiality, and topology enforcement risks.