llm-workflow-engine/llm-workflow-engine

What it solves

LLM Workflow Engine (LWE) provides a way to interact with various Large Language Models (LLMs) directly from the command line and integrate them into automated workflows. It removes the need to use a web interface, allowing users to pipe data into LLMs and use them as part of larger system operations.

How it works

LWE acts as a CLI tool and Python library that connects to LLM providers. It supports the official OpenAI ChatGPT API (including GPT-4) and other providers via a plugin architecture. Users can build complex workflows by integrating LLM calls into Ansible Playbooks, enabling the automation of tasks that require AI reasoning.

Who it’s for

Developers, system administrators, and power users who want to integrate LLM capabilities into their terminal-based workflows, Python scripts, or automation scripts using Ansible.

Highlights

• Command Line Interface: Interact with ChatGPT/GPT-4 and other LLMs directly in the shell.
• Multi-Provider Support: Compatible with OpenAI, Cohere, Huggingface, and others via plugins.
• Workflow Integration: Ability to integrate LLM calls into Ansible Playbooks for larger automation workflows.
• Extensible Architecture: Simple plugin system for adding custom functionality.
• Python API: A library for using LLMs within Python scripts.
• Tool Use: Support for tool use for compatible providers.