microsoft

microsoft / AI-Engineering-Coach

Public

better agentic engineering

38
8
89% credibility
Found May 17, 2026 at 43 stars 2x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

AI Engineer Coach is a VS Code extension that analyzes how you use AI coding assistants. It reads your local conversation logs from VS Code, GitHub Copilot, Claude, Xcode, Codex, OpenCode, and GitHub Copilot CLI, then shows you insights about your coding patterns, practice scores, and areas for improvement. The tool detects 45 common anti-patterns across five categories: prompt quality, session hygiene, code review, tool mastery, and context management. It tracks code production, token usage, and work-life balance. Everything is analyzed locally on your machine—no data is sent anywhere. The extension is an open-source community project by Microsoft employees, not an official Microsoft product.

How It Works

1
🔌 You install the extension

You add AI Engineer Coach to VS Code from the marketplace. It sits quietly in your sidebar, ready to help.

2
📊 Your coding sessions are discovered

The extension automatically finds your conversation logs from VS Code, Claude, Xcode, and other AI tools you use. Everything stays on your computer.

3
✨ You open your personal dashboard

A beautiful dashboard shows your practice scores, daily activity charts, and week-over-week trends. You see how you've been using AI coding assistants.

4
You explore different insights
⚠️
Anti-Patterns

You discover habits that might be holding you back, like vague prompts or running sessions too long.

📈
Code Production

You see how much code AI has generated for you, broken down by language and project.

🎯
Context Health

You check if your workspace instructions are helping or if context files need updating.

5
🏆 You earn achievements and level up

As you improve your AI coding habits, you unlock Bronze, Silver, Gold, and Diamond achievements. Personalized quizzes help you learn faster.

🚀 You become a better AI engineer

With clear insights and actionable tips, you write better prompts, manage context better, and get more value from your AI coding assistant.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 43 to 38 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is AI-Engineering-Coach?

AI-Engineering-Coach is a VS Code extension that analyzes your AI coding assistant usage and turns raw session logs into actionable insights. Built in TypeScript, it reads data from VS Code, GitHub Copilot for Xcode, Claude, Codex, OpenCode, and the GitHub Copilot CLI, then displays everything in a unified dashboard. You get practice scores, anti-pattern detection (45 rules across prompt quality, session hygiene, code review, tool mastery, and context management), token burndown tracking, and even gamified achievements. All analysis happens locally on your machine - nothing leaves your device.

Why is it gaining traction?

The hook is comprehensive multi-harness support. Most tools track one AI assistant; this one unifies six different tools under a single lens. The anti-pattern detection goes beyond simple metrics - it scores your "agentic readiness" by scanning for context files like .github/copilot-instructions.md or CLAUDE.md, checking hook coverage, and measuring how well you leverage tools and custom instructions. The Skill Finder actively discovers repeated prompt patterns and suggests turning them into reusable skills. It fills a gap for developers who want to improve not just what they build, but how they collaborate with AI.

Who should use this?

Senior developers and tech leads evaluating AI engineering practices across their team. Mid-level engineers wanting structured feedback on their AI usage habits. Anyone running Claude Code or GitHub Copilot CLI who wants to track token consumption and identify workflow inefficiencies. Teams with multiple developers using different AI tools who need a single visibility layer.

Verdict

At 38 stars, this is an early-stage Microsoft community project with solid test coverage and TypeScript throughout. The credibility score of 0.8999999761581421% is respectable for a niche internal tooling project. Privacy-conscious developers will appreciate the local-only analysis, but the maturity level means some features (Token Usage, Burndown) are temporarily disabled. Worth installing if you use multiple AI assistants and want quantitative insight into your engineering patterns - just don't expect polished documentation or commercial support.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.