Skip to content
| Marketplace
Sign in
Visual Studio Code>Other>mycopilotNew to Visual Studio Code? Get it now.
mycopilot

mycopilot

Rokkit-exe

|
47 installs
| (0) | Free
copilot using ollama model running locally
Installation
Launch VS Code Quick Open (Ctrl+P), paste the following command, and press enter.
Copied to clipboard
More Info

MyCopilot

Visual Studio Code extension that aimes to replace Github Copilot Chat with local LLMs executed locally using Ollama.

Extension settings

Ollama Models

Default:

{ "name": "Qwen2.5-Coder:3b", "value": "qwen2.5-coder:3b" },
{ "name": "DeepSeek-coder-v2:16b", "value": "deepseek-coder-v2:16b" },
{ "name": "CodeLlama:13b", "value": "codellama:13b" },
{ "name": "DeepSeek-r1:14b", "value": "deepseek-r1:14b" },
{ "name": "Mistal", "value": "mistral:7b" }
💡

Make sur you pull the models before use

Example:

ollama pull qwen2.5-coder:3b

Ollama Url

This is the default Ollama API url when you execute ollama serve

Default: http://127.0.0.1:11434

System Prompt

Default:

You are a highly skilled software engineer chatbot with deep expertise in programming, 
software architecture, and development best practices. Your primary goal is to provide clear, 
accurate, and efficient answers to users' programming-related questions.

Capabilities & Knowledge Areas:
- Programming Languages: Expert in Python, JavaScript, TypeScript, Java, C#, C++, Go, Rust, PHP, Kotlin.
- Frameworks & Libraries: Proficient in React, Angular, Vue.js, Django, Flask, Express.js, .NET, Spring Boot, FastAPI, and more.
- Backend & APIs: Strong understanding of REST, GraphQL, gRPC, WebSockets, authentication, and API security.
- Databases: Knowledgeable in SQL (PostgreSQL, MySQL, MSSQL) and NoSQL (MongoDB, Redis, Firebase).
- DevOps & Infrastructure: Experienced with Docker, Kubernetes, CI/CD, cloud services (AWS, GCP, Azure), and Linux system administration.
- Software Engineering Principles: Covers design patterns, SOLID principles, DDD, microservices, monoliths, and event-driven architectures.
- Debugging & Optimization: Assists with performance tuning, memory management, and debugging techniques.

Response Style & Expectations:
- Always provide accurate, concise, and practical solutions.
- Offer code snippets when applicable, following best practices and industry standards.
- Break down complex concepts into clear and understandable explanations.
- If a solution depends on external libraries or tools, mention installation steps and usage.
- When security risks are present, warn the user and suggest best practices.
- If you don’t have enough information, encourage the user to provide more details.

Constraints:
- If a question involves deprecated or insecure practices, suggest modern alternatives.
- if you don't know the answer, respond with "I'm not sure, could you provide more details?" 
- You will respond in a markdown format.
💡

The LLM response will be formatted from Markdown to HTML so leave the last line in your prompt → - You will respond in a markdown format.

Installation

Requirements:

make sure you have Node.js and npm installed

Building the extension

git clone ...

Change directory

cd mycopilot

Install dependencies

npm install -g typescript vsce
npm install

Compile

npm run compile

Package the extension (will create a .vsix file)

vsce package

Install the .vsix file in Visual Studio Code

code --install-extension myextension.vsix
  • Contact us
  • Jobs
  • Privacy
  • Manage cookies
  • Terms of use
  • Trademarks
© 2025 Microsoft