Skip to content

What Is an AI Dev Request?

DCS lets you edit text, images, and SEO yourself — but some changes need code: a new page, a layout change, a custom feature. A development request is how you ask for those. You describe what you want in plain language, an AI-assisted pipeline drafts the change, and a human reviews it before anything ships.

This page is a quick orientation. For the full walkthrough, see What are Development Requests?.

How it works

  1. You describe the change. Write what you want in plain language — for example, "Add a testimonials section to the home page with three customer quotes."
  2. It's analyzed and classified. The request is triaged: whether it can be automated, what kind of change it is, and how big a release it implies.
  3. The pipeline implements it. For automatable requests, an AI-assisted development pipeline implements the change, tracked and version-controlled through GitHub.
  4. You review it. The change is deployed to a preview environment for you to check. You approve it when it is right, or ask for adjustments.
  5. It goes live. Once approved, the change deploys to production and the request is marked complete.

Every change is human-reviewed

The pipeline is AI-assisted, not autonomous. Nothing reaches your live site until a human has reviewed it, and you get to preview and approve the result before it ships. The work flows through GitHub with full version control, so every change is tracked and reversible.

The request lifecycle

A development request moves through a clear set of states:

submittedtriagein-progressin-reviewcompleted

Two off-ramps can end a request early: rejected (declined) and archived (closed out of the active list). Requests are processed one at a time per site, so changes never collide.

What it's good for

Development requests work best for well-scoped, visual changes — adding sections, creating standard pages, adjusting colors and layout, adding components. Deeply custom or security-sensitive work (payment processing, authentication changes) is handled with human developers rather than the automated pipeline.

Next steps