Init MCP
OfficialClick on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Init MCPCreate an API that reviews source code for security issues"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Describe an endpoint in plain English. Init ships it as a live, typed, versioned HTTPS API in seconds, with auth and secrets handled. An endpoint can combine LLM reasoning, web search and scraping, document processing, image, video and voice generation, email, maps, a database, file storage, compute and scheduled jobs.
This repo holds Init's open-source clients:
Package | What it does |
| |
JavaScript client for Node.js 18+ | |
MCP server for Claude Code, Cursor, Codex, Windsurf and any MCP client |
In your coding agent
Install the agent kit, then type:
/backend email me the top Hacker News posts every morningOr add the MCP server directly:
claude mcp add init --env INIT_API_KEY=init_sk_... -- npx -y @init-computer/mcp{
"mcpServers": {
"init": {
"command": "npx",
"args": ["-y", "@init-computer/mcp"],
"env": { "INIT_API_KEY": "init_sk_..." }
}
}
}Related MCP server: Somewhere MCP
From the terminal
npm install -g @init-computer/cli
init login
init generate "An API that takes a company website and returns a sales brief with sources"
init run <route> --data website=https://www.snowflake.comGet an API key at init.computer. See the CLI reference.
From code
import { createClient } from "@init-computer/sdk";
const init = createClient(); // reads INIT_API_KEY
const api = await init.generate({
prompt: "Extract line items, totals and the due date from an invoice PDF",
});
const result = await init.run(api.url, { fileUrl: "https://example.com/invoice.pdf" });See the SDK reference.
Examples
init-computer/examples has runnable examples and starter templates. You can also generate and version endpoints in CI with init-action, or use the Vercel AI SDK and n8n integrations.
Built with Init
[](https://www.init.computer/?utm_source=badge)Development
npm install
npm test # pack checks plus CLI, SDK and MCP smoke testsThe packages are npm workspaces. Releases are tagged cli-v*, sdk-v* or mcp-v*, and CI publishes them to npm with provenance. See CONTRIBUTING.md.
License
This server cannot be deployed
Maintenance
Related MCP Connectors
Backend for AI agents: one batched tool over ~668 actions (hosting, storage, auth, payments).
Backend platform for AI agents: one agentstack.execute tool runs batched steps over +700 actions (hosting, storage, auth, payments, CRM, app builder).
Build, validate, deploy — HTTP APIs, cron jobs, webhooks and MCP tools — from your AI client.
Hosted runtime for persistent agent teams, durable workflows, memory, schedules, and goals.
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceTurns coding agents into full-stack builders by enabling backend features like auth, databases, file storage, serverless functions, and LLMs to be added to apps in seconds.8 npmApache 2.0
- AlicenseNot gradedqualityBmaintenanceEnables AI assistants to build, deploy, and inspect apps through a hosted backend service, including reading platform docs, deploying raw source, inspecting logs, verifying live pages in browsers, and managing project tasks.MIT

@manybases/mcpofficial
AlicenseNot gradedqualityCmaintenanceEnables AI agents to create and run a backend with database, auth, file storage, server-side functions, and a site, all without opening a browser.16 npmMIT- AlicenseNot gradedqualityCmaintenanceA remote MCP server that lets a coding agent turn a plain-language idea into a real, shareable product — creating the project and data model, serving live API endpoints, deploying custom endpoints and server-side functions, and reading request and function logs. It handles sign-up, OAuth-backed access, and OpenAPI spec generation so people can open, sign into, and use the app the agent builds.MIT