Orchestration Scaffold

What problem does it solve?

This Skill helps you scaffold a production-minded local AI orchestration stack by wiring together an orchestrator, a supervisor runtime, local inference, and MCP tool access with explicit safety and termination controls.

Core Features & Use Cases

• Local multi-agent orchestration blueprint: Sets up an OpenCode developer interface that connects to a LangGraph supervisor via an A2A protocol, with a clean division between planning and execution models.
• Tool access via MCP servers: Defines MCP filesystem and custom tool servers for workspace I/O and domain tooling, suitable for agent tool-calling workflows.
• Production safety harness: Provides guardrails such as pre-tool-use hooks with blocklists, enforced loop termination budgets, generator/evaluator separation to prevent self-grading, and an explicit state handoff contract.
• Operational readiness: Includes SLIs/alerts/runbook patterns plus observability components (Prometheus/Grafana/Loki) to validate task completion, latency, and error rates.

Quick Start

Use this Skill to have an AI generate a Docker Compose-based orchestration stack for your homelab and produce the required OpenCode and LangGraph integration files for multi-agent execution.