microsoft/AI-Engineering-Coach

What it solves

It helps developers improve their "agentic engineering" skills by analyzing how they interact with AI coding assistants. Instead of blindly using AI, developers can identify bad habits (anti-patterns), track their productivity, and optimize how they provide context to AI agents to get better results.

How it works

The tool operates as a VS Code extension or a canvas within the GitHub Copilot app. It reads local AI session logs on the user's machine to generate analytics without sending data to external servers. It uses a set of 45 rules to detect anti-patterns in prompt quality and session hygiene, and it can leverage the built-in VS Code Copilot language model for advanced features like skill discovery and context reviews.

Who it’s for

Software engineers who use AI coding assistants and want to move from basic prompting to professional agentic engineering by measuring their usage and refining their workflows.

Highlights

• Anti-Pattern Detection: Identifies 45 different rules across prompt quality, tool mastery, and context management to flag inefficient AI usage.
• Context Health Scoring: Audits instruction files and maps workspace context to ensure the AI agent is "ready" for the task.
• Skill Finder: Analyzes repeated prompt patterns to help users turn them into reusable skills.
• Local-First Privacy: All parsing and analytics are performed locally on the user's machine.
• Gamified Learning: Includes an XP-based achievement system and a Learning Center with personalized quizzes based on actual usage.