How to Set Up and Use Oh My Pi: A Step-by-Step Tutorial

Oh My Pi (omp) is a powerful AI coding agent that lives in your terminal. In this hands-on tutorial, we'll walk through installation, configuration, and real-world usage so you can start coding with AI assistance right away.

Prerequisites

Before you begin, make sure you have:

• macOS, Linux, or Windows with a modern terminal
• curl, Homebrew, or Bun installed (pick your preferred method)
• An API key for at least one LLM provider (OpenAI, Anthropic, DeepSeek, etc.)
• Bun ≥ 1.3.14 (for the Bun installation method)

Step 1: Install Oh My Pi

The fastest way to install on macOS or Linux is with a single command:

curl -fsSL https://omp.sh/install | sh

Alternatively, if you use Homebrew:

brew install can1357/tap/omp

For Windows users, open PowerShell as Administrator and run:

irm https://omp.sh/install.ps1 | iex

Step 2: Configure Your LLM Provider

Oh My Pi supports over 40 LLM providers. Set your API key as an environment variable or in a config file:

# For OpenAI
export OPENAI_API_KEY="sk-your-key-here"

# For Anthropic
export ANTHROPIC_API_KEY="sk-ant-your-key-here"

# For DeepSeek
export DEEPSEEK_API_KEY="your-deepseek-key"

Step 3: Set Up Shell Completions

Oh My Pi generates its own completion scripts, so they never drift from the actual CLI:

# zsh — add to ~/.zshrc
eval "$(omp completions zsh)"

# bash — add to ~/.bashrc
eval "$(omp completions bash)"

# fish
omp completions fish > ~/.config/fish/completions/omp.fish

Step 4: Your First Task

Let's put Oh My Pi to work. Navigate to a project directory and start a session:

cd ~/projects/my-app
omp

Once inside the interactive session, you can ask Oh My Pi to perform tasks directly. For example:

# Ask it to explain the codebase structure
> "What does this project do? Give me a high-level overview."

# Ask it to find and fix a bug
> "Find the bug in the authentication handler and fix it."

# Refactor code
> "Extract the database connection logic into a separate module."

Step 5: Advanced Features

Using LSP Integration

Oh My Pi integrates with your project's language server to provide code intelligence — go-to-definition, find references, hover information, and more:

> "Find all references to the User model."
> "Show me the definition of the authenticate function."

Running Python Code

Oh My Pi can execute Python code inline, making it perfect for data analysis and ML workflows:

> "Run this Python script and show me the results."
> "Analyze the CSV data in ./data/sales.csv and create a summary."

Browser Automation

You can ask Oh My Pi to interact with web pages — scrape data, fill forms, or test web applications:

> "Go to the login page and test the form validation."
> "Scrape the latest articles from this blog and save them as markdown."

Step 6: Debugging with DAP

Oh My Pi includes a full Debug Adapter Protocol (DAP) integration with 28 operations. Set breakpoints, inspect variables, and step through code:

> "Set a breakpoint at line 42 of app.py and start debugging."
> "Step into the calculate_total function and show me the variable values."

Tips for Best Results

• Use specific models for specific tasks: Fast models for simple edits, strong models for complex reasoning
• Leverage subagents: For complex tasks, Oh My Pi can spawn specialized sub-agents that work in parallel
• Keep context focused: Start fresh sessions for unrelated tasks to maintain context quality
• Use hash-anchored edits: Trust that edits are verified — Oh My Pi won't leave files in an inconsistent state

Conclusion

Oh My Pi transforms your terminal into an AI-powered development environment. With support for 40+ LLM providers, deep LSP/DAP integration, and a robust tool harness, it's one of the most capable terminal-based coding agents available today. The installation takes seconds, and the productivity gains start immediately. Try it on your next project and see the difference a true terminal AI agent makes.