Skip to content

Generator Workflow

Dockit's generation flow is split into an input page and a result page. The input view stays focused on configuration, while validation review and editing happen on a separate result screen.

1. Input Stage

The Generator combines the following inputs.

  • Presets such as docker-api, next-postgres, worker-queue, express-mongodb, and fastapi-postgres
  • Basic app data like name, domain, port, and replicas
  • Data layer choices such as database and cache
  • Delivery and advanced controls for Pro

Free users stay on a fixed development profile with a Docker Compose based flow.

2. Policy Checks

Before generation starts, Dockit verifies that the selected combination is allowed by policy.

  • Unsupported delivery targets
  • Options outside the current plan boundary
  • Incompatible database, cache, and runtime combinations

If the selection is invalid, generation is blocked immediately.

3. Generation Engine

Free

  • Local template generation
  • Fast and deterministic
  • No server call

Pro

  • Calls createGenerationJob
  • Registers work in the server-side generation job queue
  • Reuses a validated cache hit when available
  • Runs AI generation, fallback repair, and validation on cache miss

4. Result Page

After generation completes, Dockit moves to a dedicated result page. Teams can do the following there.

  • Browse the file list
  • Review the validation summary
  • Inspect bundle diff summaries
  • Edit files directly
  • Export as a ZIP

5. Delta Regeneration

Pro generation uses delta generation when a baseline bundle exists, so only the affected files are regenerated.

  • Calculates the difference between the previous and current input
  • Estimates impacted files
  • Regenerates only the files that need AI work
  • Reuses the rest from the baseline bundle

This reduces response time and avoids unnecessary token use on repeated generations.

6. Save and Reuse

  • Generated output can be saved to a project
  • Pro can reuse cached validated bundles for identical fingerprints
  • Teams can reopen the project later and continue editing or exporting
  • Projects stores reusable saved work
  • Templates lets users re-enter supported stack starting points
  • Billing controls plan access
  • Settings exposes account controls and admin tools where allowed

Dockit documentation