What is AwpakAI?

AwpakAI lets you create and run powerful workflows modeled as directed graphs. These graphs can orchestrate artificial intelligence agents, system commands, and HTTP requests.

You don’t need to be a programmer to use it:

• Developers can integrate AwpakAI as a Rust library for full control and advanced customization.
• Non‑programmers can create and execute graphs directly using the standalone command‑line tool awpak-ai-cmd-client, without writing a single line of code.

How it works

Each node in a graph can:

• Execute a system command
• Make an HTTP request (e.g., to a REST API)
• Invoke an AI agent (such as a language model or custom logic)

Graphs have an internal context where the output of every executed node is stored. Subsequent nodes can access outputs from previously executed nodes, enabling flexible and dynamic data flows.

Use Cases

AwpakAI is ideal for:

• Coordinating language models (LLMs) or AI agents in complex workflows
• Automating pipelines combining AI, web services, and shell commands
• Defining flexible, declarative flows in JSON for easy maintenance and dynamic reconfiguration

Creating Graphs Easily

Writing JSON graph definitions manually can be complex due to the available options. To make this easier, we provide a web-based editor to build graphs interactively and export them as JSON.

The Getting Started guide shows how to create and customize a graph using the web editor, and how to run it either from your Rust project or directly with awpak-ai-cmd-client.