Track the generation process, view activity logs, and monitor the status of your documentation projects.
Monitoring project activity in AI Docs provides crucial visibility into the documentation generation process, ensuring your projects are up-to-date and functioning as expected. This feature allows you to track the real-time status of your documentation pipeline, review generation reports, and access key project metadata. Regularly checking project activity helps you identify and address any issues promptly, ensuring continuous, high-quality documentation for your repositories.
To monitor the activity of a specific project:
This tab provides a detailed, real-time view of the documentation generation pipeline and a summary of the latest generation.
The core of project activity monitoring is the documentation generation pipeline. This multi-step process transforms your GitHub repository into comprehensive documentation. The Activity tab displays a timeline of these steps, indicating their current status and any associated details.
The activity dashboard automatically polls for updates every 3 seconds when the pipeline is running, providing real-time progress without requiring a page refresh.
The documentation generation process follows a structured flow, from connecting your repository to publishing the final documentation.
Each step in the pipeline progresses through different statuses:
Here's a breakdown of each pipeline step:
This initial step confirms that your GitHub repository has been successfully linked to AI Docs. It displays basic repository information, such as the owner, repository name, and default branch. This step is always marked as completed once the project is established.
During this step, AI Docs connects to your GitHub repository to download and ingest its files. This involves pulling the latest code from the specified default branch.
For more details on how AI Docs interacts with GitHub, refer to GitHub Integration.
Once the repository files are ingested, AI Docs begins analyzing your codebase. This step extracts code intelligence, identifies key components, functions, and their relationships, laying the groundwork for intelligent documentation generation.
This crucial step involves creating vector embeddings for your code chunks. These embeddings are numerical representations of your code that capture its semantic meaning, enabling efficient retrieval and contextual understanding by the AI model.
To learn more about how embeddings are used for semantic search and context, see Vector Search with Qdrant and Deep Dive into Search & RAG.
With the code analyzed and embedded, the AI model plans the optimal documentation hierarchy and structure. This involves determining the logical organization of pages, sections, and topics based on the codebase's architecture.
For a deeper understanding of how AI structures content, refer to Documentation Structure and Display.
In the final generation step, the AI model writes the content for each documentation page based on the planned structure and the analyzed code. This is where the comprehensive, visually appealing documentation is generated.
For more information on the AI architecture and content generation, refer to AI-Powered Generation.
Once the documentation pipeline completes, the Activity tab presents a "Generation Report" summary. This report provides key statistics about the recently completed documentation run.
Below these metrics, you'll find an additional insight: the ratio of documentation pages generated per source file, offering a quick overview of the documentation's density relative to your codebase.
The Activity tab also includes a sidebar displaying essential project metadata: