Skip to content

AI-native cloud platform

Agents operate. You control.

Agents now write the software. Guidal is where they deploy and run it, with people in control.

▸ orders-api/

  • ├ main.py
  • ├ requirements.txt
  • └ Dockerfile

$ guidal deploy ./orders-api

$ guidal services create --name orders-api --image <image>

Illustrative

  1. 01DetectDockerfile or auto-detected runtime
  2. 02Buildinto a container image
  3. 03Pushstored with your project
  4. 04Runrolling deploy over HTTPS

200 · live

orders-api-shop.guidal.cloud

HTTPS · certificate issued automatically

Illustrative

Agents now do the work, on clouds built for people.

  • Made for agents

    Your agent works through Guidal's own tools, not a console and keys built for people. Fewer wasted steps.

  • Less DevOps for small teams

    Deploys, backups, logs and failure diagnoses come with the platform. Your agent handles the routine; you stay in charge.

  • Safe for agents that never stop

    Loops and multi-agent setups can build, deploy and fix around the clock. Guidal enforces limits at the platform level. Guidal's agent tools can't delete your data or change your billing.

Illustrative

Your agent ships it and fixes it.

Your agent works from your IDE, a terminal or a server. It creates, builds, deploys and checks on your apps.

Guidal
orders / production
api
Live
api-orders.guidal.cloud
DATABASE_URL
postgres
Ready
orders-db · backups on
Cursor · your AI IDE

Deploy my FastAPI service with a Postgres database and give me a public HTTPS URL.

Plan: one web service + managed Postgres, on Guidal.

$

guidal deploy ./orders-api

Built from your folder. Starting api from it.

GET /health and /docs: both 200 over HTTPS.

Live. DATABASE_URL is injected; no secrets in the image.

What Guidal did

Created project orders

ready

Provisioned Postgres orders-db

backups on

Built and rolled out api

live on HTTPS

Linked orders-db to api

DATABASE_URL injected

Confirmed api

running · 1 copy

Console, CLI or agent.

Not using an agent? You do not need one. Deploy, follow logs and metrics, and manage databases and secrets from the console or your terminal. Your agent, when you use one, calls the same operations.

  • Console

    Deploy an image, scale it, and follow its deployments, logs and metrics in the browser.

  • CLI

    Build from source, read logs, roll back and open a database session from your terminal.

  • Agent

    Works with Claude Code, Cursor and other MCP clients.

If it runs in a container, it runs here.

  • Web services

    Any HTTP app, any language. Rolling deploys and an HTTPS hostname.
  • Workers

     Planned
    Long-running background processes and queue consumers.
  • Scheduled jobs

     Planned
    Cron on your schedule, with a history of every run.
  • Postgres

     Beta
    Managed, with backups, restore to a new database and deletion protection.
  • Redis

     Beta
    Managed, password-protected, private to your project.
  • Object storage

    S3-compatible buckets for files and uploads.

Built into every project

  • HTTPS hostname · custom domains on the planned paid plans
  • Secrets injected at runtime
  • Private networking

Your agent keeps it running.

Ask your agent to check on your app. It reads your logs and metrics, finds the likely cause and fixes what it can.

Your agent · shop
Why is checkout-api returning 500s?

Checked logs and metrics

status
Running · 2 of 2 ready
errors in the last hour
214 · cannot reach the database
memory
61% of limit

AgentThe app cannot reach its database. orders-db is no longer linked to checkout-api; I can link it again.

Example output. Your agent does this only when you ask.

Start with a folder.

Start with a source folder or an image. Guidal runs it and gives you a URL.