Let AI generate coding practices & discussion topics

In this section, you'll see how to use the Git integrations & the IDE plugin to generate coding practices, and how to analyze the results

Run AI through Git integrations

How does it work?

With a continuous scan of your code changes in Git repositories, Packmind AI suggests new coding practices and raises discussion topics for your practice review workshops. This generation engine brings innovative support for Tech leads & developers, and should be considered as an external eye capable of making suggestions and asking questions about your coding standards or why such decisions were made.
Concretely, Packmind will parse the latest commits, and from the added lines of code, it will suggest content for the next practice review workshop that the drivers must first validate.

Requirement: Git integration

Follow this dedicated guide on connecting Packmind with your Git repositories (GitLab, GitHub, Azure DevOps or Bitbucket) through REST APIs.

Configure scan and generation frequency

To tune generative AI, go to Administration** -> **AI Settings section, and identify the following sections:
Here, you can tune the configuration for each space (choose the selected space from the list in the top-right section).
Then you can configure:
  • The frequency: how often Git scans should be performed?
  • The number of positive examples of best coding practices
  • The number of negative examples of best coding practices
  • The number of questions and discussion topics
Be aware that Packmind tries to prioritize the pieces of code from which content will be generated.
In case there are too few changes, we can't guarantee you'll get generated content.

Run manual AI generation from Git repositories analysis

To run an AI analysis on your Git repositories, go to Administration** -> **AI Settings section and select a space with an active Git Configuration:
Spaces with a Git repository are marked by the icon of their provider
Then, select a time frame, meaning that Packmind AI will focus on Git commits that occurred in the chosen time frame (for instance, commits from the last 24 hours or since one week).
Select a time frame and let the AI generate practice and questions for you
Click on Trigger AI Now to start a scan.

Using your IDE

Note that the AI feature is available only for Packmind VSCode and JetBrains extensions.
The workflow is straightforward:
  1. 1.
    You select one or several files you'd like to send to Packmind from the IDE extensions.
  2. 2.
    The AI engine will analyze the code to generate up to 3 coding practice suggestions (and a correction) and up to 3 discussion topics.
Right-Click on the code editor and select the Generate Practices entry. Select only the current opened file or all the opened tabs, choose your destination space in Packmind, and you're done!
A notification will prompt to indicate the process has started.

Review suggestions and push them for the following workshops

We recommend setting notifications to be informed when Packmind AI has generated new content