agent-mesh-mcp

What problem does it solve?

It helps you manage a homelab AI control plane where multiple agents, model routers, and cluster services must coordinate without accidental coupling or runaway spend.

Core Features & Use Cases

• Agent mesh architecture: Describes how OpenClaw, Hermes, and Kubernetes MCP services should delegate tasks across the cluster.
• Cost-aware routing: Uses Prometheus and LiteLLM metrics to choose economical models and track token spend before expensive operations.
• Granular deployments: Supports separated Ansible and Makefile workflows for updating LiteLLM, Hermes, and shared services independently.
• Use case: Use it when you need to decide whether to route a request to Hermes, check budget pressure, or roll out only one AI component without restarting the whole stack.

Quick Start

Ask for an explanation of the agent mesh, the current routing rules, and the safest deploy command for the component you want to update.