How to Build an Automated Software Factory from Slack to GitHub

Start the workflow in a public Slack channel by tagging an AI factory bot (e.g., '@Wilson') with a detailed request, including instructions and reference images. This makes all work observable to the team.

The AI agent acts as a project manager, automatically triaging the request and creating a corresponding issue in your project management tool (e.g., Linear) to ensure all work is tracked.

After creating the ticket, the agent creates a new branch and a draft pull request in GitHub, then begins implementing the code changes based on the initial request.

Once the code is written, the agent performs a 'computer use verification.' It runs the new feature in a test environment and records a video of it working, including keystrokes and visual output.

The agent attaches the QA verification video to the GitHub pull request, providing immediate, visual evidence that the feature works as intended and is ready for human review.