spawnagents

Spawn and control AI agents programmatically.

Full release coming soon — spawnagents.dev

What is SpawnAgents?

SpawnAgents is a platform and SDK for creating, deploying, and orchestrating AI agents without managing infrastructure. You define what the agent should do; SpawnAgents handles the runtime, memory, tool execution, and lifecycle.

Each agent is a persistent, stateful process — not a one-shot LLM call. Agents can browse the web, write and run code, call APIs, send messages, and coordinate with other agents. You interact with them over a simple API or stream their output in real time.

Core concepts

Spawn an agent

import { spawnagents } from 'spawnagents'

const agent = await spawnagents.spawn({
  name: 'researcher',
  model: 'claude-opus-4-6',
  instructions: 'You are a research assistant. When given a topic, find the latest information and summarise it.',
})

Run a task

const result = await agent.run('Summarise the latest news on open-source LLMs')

console.log(result.output)
// → "Here is a summary of the latest developments in open-source LLMs..."

Stream output in real time

const stream = agent.stream('Write a full business plan for a SaaS analytics tool')

for await (const chunk of stream) {
  process.stdout.write(chunk.text)
}

Give agents tools

const agent = await spawnagents.spawn({
  name: 'dev-agent',
  model: 'claude-opus-4-6',
  instructions: 'You are a senior software engineer.',
  tools: ['code_execution', 'web_search', 'file_system'],
})

Persistent memory

Agents remember context across sessions. Each agent has its own isolated workspace that persists between calls — files, state, conversation history.

// First session
await agent.run('Read the file data.csv and build a summary')

// Later — same agent, same context
await agent.run('Now generate a chart from that data')

Multi-agent orchestration

const planner = await spawnagents.spawn({ name: 'planner', ... })
const executor = await spawnagents.spawn({ name: 'executor', ... })

const plan = await planner.run('Break down this project into tasks: ...')
const result = await executor.run(plan.output)

MCP support

SpawnAgents agents are compatible with the Model Context Protocol. Connect any MCP server to extend what your agents can do.

const agent = await spawnagents.spawn({
  name: 'power-agent',
  mcp: ['https://mcp.yourservice.dev/sse'],
})

REST API

Every agent is also accessible over HTTP with a stable endpoint — useful for webhooks, Zapier, n8n, or any no-code platform.

POST https://api.spawnagents.dev/v1/agents/{id}/run
Authorization: Bearer sk-sa-...
Content-Type: application/json

{ "input": "Analyse this CSV and return insights" }

Designed for developers and teams

• Instant setup — no Docker, no VMs, no GPU provisioning
• Bring your own model — Claude, GPT-4o, Gemini, open-source models
• Team access — invite teammates, share agents, manage permissions
• Audit logs — full trace of every agent action
• SDK + REST — Node.js SDK with full TypeScript support; REST API for everything else

Roadmap

• [ ] Node.js SDK (spawnagents)
• [ ] REST API + streaming
• [ ] Persistent agent workspaces
• [ ] Built-in tools (web, code, files, email, calendar)
• [ ] MCP server compatibility
• [ ] Multi-agent orchestration
• [ ] Python SDK
• [ ] Dashboard at spawnagents.dev

Stay updated

• Website: spawnagents.dev
• GitHub: github.com/spawnagents
• npm: npmjs.com/package/spawnagents

License

MIT