klyro 0.1.0

Klyro is AI pair programming in your terminal

🧠 Klyro

AI Pair Programming in Your Terminal

---

Klyro lets you pair program with LLMs directly from your terminal. Start a new project or work on your existing codebase with an AI that can read, understand, and edit your code.

⚡ Quick Install

pip install klyro

Or install from source:

git clone https://github.com/RavindraTirlangi/Klyro.git
cd Klyro
pip install -e .

🚀 Usage

# Start Klyro in your project directory
klyro

# Use a specific model
klyro --model gpt-4o

# Use a local Ollama model (auto-detected if running)
klyro --model ollama/llama3.2

✨ Features

Smart Model Detection

Klyro auto-detects your environment:

• Local Ollama running? Uses it automatically — no API key needed.
• Cloud API keys set? Picks the best available model.
• No keys at all? Offers one-click OpenRouter setup.

Slash Commands

Command	Description
/stats	Show model name, last message cost, and total session cost
/web <url>	Fetch a web page and add its content to the chat context
/export [filename]	Export chat history to a markdown file
/tokens	Report token usage for the current chat context
/add <file>	Add a file to the chat
/drop <file>	Remove a file from the chat
/diff	Show the diff of changes made
/undo	Undo the last change
/help	Show all available commands

Code Editing

• Whole file and diff-based editing modes
• Automatic linting after every edit
• Built-in git integration with auto-commits

Repository Intelligence

• Builds a repo map using tree-sitter AST analysis
• Understands your codebase structure and relationships
• Smart context management to stay within token limits

🔧 Configuration

Create a .klyro.conf.yml in your project root:

model: ollama/llama3.2
dark-mode: true
auto-commits: true

Or use environment variables:

export OPENAI_API_KEY=sk-...
export ANTHROPIC_API_KEY=sk-ant-...
export OLLAMA_API_BASE=http://localhost:11434

📦 Publishing

This project uses GitHub Actions for automated PyPI publishing. To release:

git tag v1.0.0
git push origin v1.0.0

The workflow will automatically build and publish to PyPI.

📄 License

Apache License 2.0 — see LICENSE for details.