Skip to content
DocuBot documentation

How it works

DocuBot transforms your GitHub repositories into structured, high-performance help centers. By combining advanced AI analysis with a specialized documentation framework, the system extracts knowledge directly from your code and publishes it as a static site tailored to your specific audiences.

The documentation pipeline

The journey from source code to a published help center follows a streamlined path managed through the DocuBot dashboard. You begin by defining a unique URL slug, which serves as the permanent home for your documentation.

Once you connect one or more GitHub repositories, DocuBot acts as an orchestrator. It scans the connected sources, identifies their roles—such as a primary web application or a supporting API—and prepares the content for analysis. The final output is a fully rendered static site, hosted at your chosen slug, ready for your users and support teams to access.

Context-aware AI analysis

At the core of DocuBot is an AI-driven analysis engine powered by Gemini. Unlike simple text generators, DocuBot performs a comprehensive scan of your entire codebase to understand the product’s purpose, key features, and primary use cases.

The analysis engine performs several critical tasks:

  • Information extraction: It identifies modules, setup steps, and configuration options directly from the logic of your code.
  • Deduplication: If your project spans multiple repositories, the system consolidates overlapping information to ensure the documentation remains concise and accurate.
  • Canonical summarization: The system builds a “source of truth” summary that ensures every generated page—from a quick start guide to a technical reference—remains factually consistent.

The style and audience framework

DocuBot uses a layered style model to ensure your documentation sounds professional and serves the right people. You control this through three distinct layers:

  1. Target audiences: You select who the documentation is for (such as End Users, Administrators, or Developers). This selection automatically determines which document types are most relevant to your project.
  2. General styles: You choose the overall voice of the site. Options like “DocuBot Default,” “Precision Technical,” or “Approachable Guided” dictate the tone and focus of every sentence.
  3. Documentation quadrants: Every page follows the Diátaxis framework, which categorizes content into Tutorials, How-to guides, Explanations, or References. This ensures that a tutorial feels like a lesson, while a reference page remains a neutral fact sheet.

Static site architecture

DocuBot publishes documentation as a static site using the Astro and Starlight engines. This architectural choice provides several benefits for your documentation:

  • SEO optimization: Every page has a real, permanent URL that search engines can index, making it easier for your users to find answers via Google or Bing.
  • Speed and reliability: Because the pages are pre-rendered HTML, they load instantly and do not rely on a database or complex runtime to display content.
  • Linkable content: You can share direct links to specific sections or pages with confidence, knowing the routes will remain stable.

Synchronization and change detection

To keep your documentation aligned with your latest releases, DocuBot uses a synchronization system. You can configure an automated schedule—daily, weekly, or monthly—to refresh your docs at a specific local time.

When a sync job runs, DocuBot checks for changes in your primary repository. If new code is detected, the system regenerates the documentation to reflect the updates. You can also trigger a manual synchronization at any time from the dashboard if you need to publish an urgent update immediately. The dashboard provides real-time status tracking so you can monitor the progress of every generation job.

API reference processing

If your repositories contain OpenAPI or Swagger specifications, DocuBot automatically detects them and creates interactive API references.

Instead of static text, these specifications are transformed into a dedicated API viewer powered by Scalar. This viewer allows developers to explore endpoints, view request and response schemas, and understand authentication requirements in a modern, interactive interface. You can include multiple API specifications within a single documentation project, and they will appear organized within your site’s navigation.

Support and assistance

If you encounter issues with the documentation generation process or need help configuring your project, you can reach out to the support team: