Maige

What is Maige?

Maige is an open-source infrastructure that runs natural language workflows directly on your codebase. Powered by GPT, it classifies and labels issues in GitHub repositories: it labels new issues as they arrive, labels existing issues when you comment “Maige label this,” and follows custom instructions for how issues should be classified. Companies including Highlight.io, PrecedentCal.com, and Trigger.dev use it. A free trial is available; details on premium features or paid plans are not readily available.

Who created Maige?

Ted Spare created Maige as a tool for classifying and labeling software project issues. Beyond labeling, it supports auto-assignment, auto-comments, custom instructions, code review, and code generation, available under different pricing structures.

What is Maige used for?

• Auto-labels new GitHub issues as they arrive and classifies existing ones
• Assigns tasks and adds comments on issues and pull requests based on rules you set
• Reviews incoming pull requests, including against your CONTRIBUTING.md guidelines
• Runs simple code snippets based on predefined rules
• Automatically sets up a code sandbox environment for testing
• Monitors workflows and provides feedback through a dedicated dashboard
• Customizes workflows with simple text-based rules, operating via the GitHub API

Example rules include labeling UI-related issues to specific users or tagging .env pull requests with “needs-approval.”

Who is Maige for?

Maige suits software developers, software engineers, and anyone who uses GitHub.

Getting started

Connecting a repo sets up a webhook, creates codebase embeddings, and prepares a sandbox environment. From there, write text rules describing what Maige should do with issues and pull requests — labeling, assigning, commenting, reviewing, or running code snippets. Monitor how workflows run and provide feedback through the dashboard.