Repolode
Submit
microsoft

microsoft/AI-Engineering-Coach

A VS Code extension and Copilot canvas that analyzes your local AI coding assistant session logs into practice scores, anti-patterns and output metrics.

View on GitHub
Stars
4,342
Forks
601
Open issues
56
Last commit
2 days ago

What is it?

AI Engineer Coach reads the session logs kept by your local AI coding assistants and turns them into a dashboard. It tracks practice scores and weekly trends, shows daily activity and work-hour patterns, and measures how much code is generated by language, workspace, model and harness. All parsing runs on your machine and the extension only reads session files.

The project includes 45 editable markdown rules that flag anti-patterns across prompt quality, session hygiene, code review, tool use and context management. It also finds repeated prompts that can become reusable skills, and audits instruction files and workspace context for agentic readiness. A rule editor and playground let you tune detection rules against your own data.

The dashboard ships as a VS Code extension that you build from source as a .vsix file, since it is not published to a marketplace. The same web UI can also run as a canvas inside the GitHub Copilot app, where features that depend on the local VS Code language model are hidden.

Key features

  • 01Practice scores with week-over-week trends and daily activity charts
  • 0245 anti-pattern rules covering prompts, sessions, code review and context
  • 03Generated code volume broken down by language, workspace, model and harness
  • 04Repeated prompt discovery to suggest reusable skills
  • 05Context health checks, instruction-file audits and workspace context maps
  • 06Editable detection rules with a visual editor and REPL playground

Use cases

  • Reviewing how you use AI coding assistants and spotting habits to improve
  • Auditing instruction files and project context before running agents on a codebase
  • Measuring AI-generated code volume across projects and models
  • Finding repeated prompts worth turning into shared skills

Topics

#ai-coding#coding-assistant#vscode-extension#developer-productivity#analytics#llm#copilot

FAQ

What is AI-Engineering-Coach?
A VS Code extension and Copilot canvas that analyzes your local AI coding assistant session logs into practice scores, anti-patterns and output metrics. AI Engineer Coach reads the session logs kept by your local AI coding assistants and turns them into a dashboard. It tracks practice scores and weekly trends, shows daily activity and work-hour patterns, and measures how much code is generated by language, workspace, model and harness. All parsing runs on your machine and the extension only reads session files.
Is AI-Engineering-Coach free to use? What license does it have?
AI-Engineering-Coach is open source under the MIT license. Check the license terms before using it in commercial products.
Is AI-Engineering-Coach still maintained?
Actively maintained. The last commit to AI-Engineering-Coach was on Oct 16, 2026.
What can I use AI-Engineering-Coach for?
Reviewing how you use AI coding assistants and spotting habits to improve Auditing instruction files and project context before running agents on a codebase Measuring AI-generated code volume across projects and models Finding repeated prompts worth turning into shared skills
Where can I get AI-Engineering-Coach?
The source code and installation instructions are on GitHub: https://github.com/microsoft/AI-Engineering-Coach

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects