AI-powered multi-cloud deployments from a single CLI.

Agent Cloud uses intelligent agents to analyze your project, generate a cloud-specific deployment plan, and run the real commands for AWS, GCP, or Azure.

Understand it in under 2 minutes

The goal: make deployments feel like a guided, reviewable workflow instead of a pile of one-off scripts.

One command to start

Choose your cloud provider, review the plan, then approve. Use --yes to auto-approve.

cloud-agent deploy --cloud aws

# Review plan
# Approve when ready

# Or auto-approve
cloud-agent deploy --cloud aws --yes

Simple mental model

CLI triggers a workflow. Agents do the reasoning. Providers do the real cloud work.

CLI
  -> Agents (analyze + plan + validate)
  -> Providers (AWS / GCP / Azure)
  -> Deployment URL + logs

What you get

Enough structure for production deployments, without the enterprise overhead.

Project analysis

Detect stack, runtime, dependencies, and deployment shape (static site, API, container).

Plan you can review

Deployment steps, commands, and cost estimates before anything runs.

Human-in-the-loop

Approval gates so you can pause, inspect, and then proceed (or stop).

Multi-cloud support

AWS, GCP, and Azure with cloud-specific targets and best practices.

Environment validation

Checks credentials, CLIs, permissions, and network connectivity before deployments.

Logging and history

Session-based logs, workflow history, and simple analytics for what happened and when.

Multi-cloud targets (MVP)

Agent Cloud supports common deployment targets across the three major providers.

Cloud Container Serverless Static sites PaaS
AWS ECS Fargate Lambda S3 -
GCP Cloud Run Cloud Functions Firebase + Storage App Engine
Azure Container Apps Functions Static Web Apps + Blob App Service

Next steps

Start with Getting started, or jump into Contributing.