andrewyng/aisuite ? reverse-engineered prompt

Reverse engineered prompt

Build me a small Python library that makes it easy to talk to different AI providers through one simple interface.

I want a unified chat API where I can switch between providers like OpenAI, Anthropic, Google, Ollama, and other compatible services by changing the model string, while keeping the request and response format consistent. It should support normal chat messages, streaming responses, and basic settings like temperature, token limits, and tools.

Please also add a simple agents layer on top, so I can pass plain Python functions as tools, let the model call them, and optionally run multi turn loops automatically. It would be great if it included a few ready made tool helpers for things like files, git, and shell, plus support for MCP style tools if that fits cleanly.

Keep it lightweight, easy to install, and documented with a few examples and a quickstart that shows how to use it with real provider keys or local models. If you need to check current provider docs online, go ahead.

Are you gonna build this?

make sure you review the code using coderabbit

Try freeSponsored — opens CodeRabbit in a new tab