Documentation

Everything you need to use CanAI effectively. Honest, practical, no fluff.

Quick Start

5 minutes to your first AI-assisted workflow.

1

Install

Python 3.10+ required. One command, zero mandatory dependencies.

pip install canai

Installs all 21 tools as CLI commands. Disk usage: ~5 MB.

2

Bootstrap your project

Sets up .ai/ directory, CLAUDE.md, AGENTS.md — everything AI needs to understand your code.

cd your-project/
scaffoldai . --language Python --framework FastAPI

Works with Python, JavaScript, TypeScript, Rust, Go, and 15+ other languages.

3

Generate AI context

Scans your entire codebase and creates a structured context file AI can read.

ctxbuilder . > CLAUDE.md

10,000 files: ~2 seconds. Respects .gitignore automatically.

4

Set up your API keys

Create a .env file or use the built-in encrypted vault. You need at least one AI provider key.

# .env
ANTHROPIC_API_KEY=sk-ant-...
OPENAI_API_KEY=sk-...
# Or: use the encrypted vault via Settings page
5

Start coding with AI

Chat with any LLM provider using a single interface. Switch providers without changing code.

apibridge "Explain this codebase to me" --provider claude
# Or interactive terminal pairing:
cancraft --dir src/