Skip to main content
Use this when you are starting from an empty or early repository. Your coding agent writes the app, and Monk plans and runs the infrastructure for it.

​
Before you start

  • Monk is installed and you are signed in. See Install.
  • A repository open in your coding agent, even if it only has a first commit.

​
Ask Monk

Once there is something to run, ask for a plan:
If the app needs a database or a cache, say so. Your coding agent adds the code that uses it, and Monk adds it to the plan:
Try it on your own machine before you pay for a cloud:
When it works, take it to a cloud:

​
What you review

Each deploy opens a plan in the local dashboard: the services, their connections, missing configuration and, for cloud deploys, the machines and estimated cost. Credential forms open there too. Nothing is created until you approve.

​
What you get

  • The app running locally or in your cloud account
  • Dockerfiles and a MANIFEST in your repository, written by your coding agent and validated by Monk
  • A model of the app that later requests build on
Commit the Dockerfiles and MANIFEST with your code. They describe how the app runs.

​
Keep going

As the app grows, ask for what it needs:

​
Next