Gwenflow — AI skill for Claude Code

Why Gwenflow?

Gwenflow, a framework designed by [Gwenlake](https://gwenlake.com), streamlines the creation of customized, production-ready applications built around Agents and Large Language Models (LLMs). It provides developers with the tools to integrate LLMs and Agents into efficient, scalable solutions.

**Key capabilities:**

• Multiple LLM providers — OpenAI, Anthropic, Azure, Mistral, Google, Ollama, DeepSeek
• Autonomous agents — agentic loop with tool use, memory, and structured output
• Multi-agent flows — DAG-based pipelines defined in code or YAML
• RAG pipeline — document readers, vector stores, and retrieval-augmented generation
• Streaming — sync and async streaming for all providers
• Telemetry — built-in OpenTelemetry tracing

Installation

pip install gwenflow

Install the latest from the main branch:

pip install -U git+https://github.com/gwenlake/gwenflow.git@main

Quick Start

from gwenflow import ChatOpenAI

llm = ChatOpenAI(model="gpt-5-mini")
response = llm.invoke("Describe Argentina in one sentence.")
print(response.text)

Multiple LLM Providers

Swap providers with a single import — the agent API is identical across all of them.

from gwenflow import (
    ChatOpenAI, ChatAnthropic, ChatGoogle, ChatMistral,
    ChatDeepSeek, ChatOllama, ChatAzureOpenAI, ChatGwenlake,
)

openai_llm    = ChatOpenAI(model="gpt-5-mini")
anthropic_llm = ChatAnthropic(model="claude-sonnet-4-6")
google_llm    = ChatGoogle