Open source CLI tool

Deploy to any cloud
with one command

AI agents analyze your project, generate deployment plans, and ship to AWS, GCP, or Azure. You approve before anything runs.

$npm install -g agent-cloud
terminal

Everything you need to ship

From project analysis to production deployment, agent-cloud handles the complexity so you can focus on building.

AI-Powered Analysis

Three specialized agents analyze your project, plan deployments, and validate your environment.

Multi-Cloud Support

Deploy to AWS, Google Cloud, or Azure. One CLI, three clouds. Switch providers with a single flag.

Human-in-the-Loop

Review AI-generated plans before execution. Approve, modify, or reject. You stay in control.

CLI-First Design

Built with Commander.js. Interactive prompts, spinners, and clean terminal output.

Security Hardened

Shell injection prevention on every exec call. Command allowlists and path traversal protection.

Zero-Config Deploy

Auto-detects runtime, framework, and dependencies. Generates Dockerfiles for Node.js, Python, and Go.

Dockerfile Generation

No Dockerfile? Automatically generates optimized multi-stage builds for your stack.

Cost Estimation

Get estimated monthly costs before deploying. Compare pricing across all three providers.

Deployment History

Track every deployment with full history. View URLs, resources, costs, and durations.

Credential Safety

Uses your existing cloud CLI credentials. No tokens stored. Works with SSO and MFA.

Dry Run Mode

Preview the entire plan without creating resources. Perfect for CI/CD pipelines.

Easy Cleanup

Remove all deployed resources with a single command. Select specific deployments to tear down.

How it works

Four phases from code to cloud. AI handles the complexity, you keep the control.

01

Analyze

AI agents scan your project, detect your tech stack, framework, databases, and dependencies. Works with Node.js, Python, Go, and static sites.

cloud-agent analyze
02

Plan

The deployment planner maps your project to optimal cloud services, estimates costs, and generates infrastructure commands.

cloud-agent deploy --cloud aws
03

Approve

Review the generated plan with services, estimated costs, and commands. Approve to proceed, or use --dry-run to preview first.

? Proceed with deployment? (y/N)
04

Deploy

Real cloud resources are created. ECS clusters, Cloud Run services, Container Apps. Get a live URL when it completes.

URL: https://my-app.us-east-1.aws.com

One CLI, three clouds

Same commands, same workflow. Switch providers with a single flag. No vendor lock-in.

AWS

Amazon Web Services

Deploy containerized apps to ECS, serverless functions to Lambda, and static sites to S3 with CloudFront CDN.

Services
ECS FargateLambdaS3 + CloudFrontCloudWatchIAMVPC
$ cloud-agent deploy --cloud aws
GCP

Google Cloud Platform

Container deployments on Cloud Run, serverless with Cloud Functions, and static hosting via Firebase.

Services
Cloud RunCloud FunctionsFirebase HostingApp EngineCloud StorageCloud Build
$ cloud-agent deploy --cloud gcp
Azure

Microsoft Azure

Modern container apps, serverless functions, static web apps, and PaaS deployments with automatic resource groups.

Services
Container AppsAzure FunctionsStatic Web AppsApp ServiceBlob StorageResource Groups
$ cloud-agent deploy --cloud azure

CLI commands

Every command you need from init to cleanup. Built with Commander.js for a polished terminal experience.

init

Set up API keys, cloud CLI detection, and preferences

$ cloud-agent init
analyze

Analyze your project and get cloud recommendations

$ cloud-agent analyze
--ai--local--path <dir>
deploy

Deploy your application to any cloud provider

$ cloud-agent deploy --cloud aws
--cloud <provider>--yes--app-name <name>--port <number>--image <image>--build-dir <dir>--dry-run
status

Check if your environment is ready for deployment

$ cloud-agent status --cloud gcp
--cloud <provider>
logs

View logs from your deployed services

$ cloud-agent logs --cloud aws --service my-app
--cloud <provider>--service <name>--lines <n>
cleanup

Remove previously deployed cloud resources

$ cloud-agent cleanup
--cloud <provider>
history

Show your deployment history and stats

$ cloud-agent history

Architecture

Three clean layers: AI agents for intelligence, a workflow engine for orchestration, and cloud providers for execution.

AI Agents

Mastra Framework

Analyzer Agent

Scans projects, detects stack

Deployment Agent

Plans infrastructure, estimates costs

Validator Agent

Checks CLI, auth, permissions

Workflow Engine

5-Phase Pipeline

Validate

Environment readiness check

Analyze + Plan

Project scan and service mapping

Approve + Deploy

Human gate then real execution

Cloud Providers

Real Infrastructure

AWS Provider

ECS, Lambda, S3, CloudFront

GCP Provider

Cloud Run, Functions, Firebase

Azure Provider

Container Apps, Functions, Static

Get started

Four commands from install to production.

1
Install
$ npm install -g agent-cloud
2
Configure
$ cloud-agent init
3
Analyze
$ cloud-agent analyze
4
Deploy
$ cloud-agent deploy --cloud aws