Multi-Agent System

Multi-agent architecture with A2A protocol peers, MCP tool servers, and an orchestrating agent. Agents discover each other, delegate tasks, and share tools.

Name	`multi-agent`
Version	`0.1.0`
Tags	ai, agents, a2a, mcp, multi-agent
Source	`composites/multi-agent.yaml`

Render it

# CLI
npx @nanohype/sdk render-composite multi-agent --out ./my-app

# SDK
import { LocalSource, renderComposite } from "@nanohype/sdk";
const result = await renderComposite(source, "multi-agent", variables);

# MCP — from an agent
get_composite({ name: "multi-agent" })

What it draws together

Template	Rendered at	Notes
`monorepo`	root	the root project
`a2a-agent`	`packages/orchestrator`
`agentic-loop`	`packages/agent`
`mcp-server-ts`	`packages/tools`
`guardrails`	`packages/guardrails`	only when `IncludeGuardrails`
`eval-harness`	`packages/evals`
`ts-service`	`apps/api`
`module-auth-ts`	`packages/auth`
`k8s-deploy`	`infra`

Variables

Required

ProjectName string: Project name Must be lowercase kebab-case.

Optional

LlmProvider string — defaults to anthropic: LLM provider for agent reasoning
IncludeGuardrails bool — defaults to true: Include safety guardrails