Skip to content

Welcome to aipartnerupflow¶

aipartnerupflow is a unified framework for orchestrating and executing tasks across multiple execution methods. It manages when tasks run, how they depend on each other, and ensures everything executes in the right order—whether you're calling HTTP APIs, executing SSH commands, running Docker containers, or coordinating AI agents.

Problems We Solve¶

Are you struggling with these common challenges?

  • Complex Task Dependencies

    Manually tracking dependencies, ensuring proper execution order, and handling failures across complex workflows becomes a nightmare. You end up writing custom coordination code and dealing with race conditions.

  • Multiple Execution Methods

    You need to call HTTP APIs, execute SSH commands, run Docker containers, communicate via gRPC, and coordinate AI agents—but each requires different libraries and integration patterns.

  • Traditional Tasks + AI Agents

    You want to add AI capabilities to existing workflows, but most solutions force you to choose: either traditional task execution OR AI agents. You're stuck with all-or-nothing decisions.

  • State Persistence & Recovery

    When workflows fail or get interrupted, you lose progress. Implementing retry logic, checkpointing, and state recovery requires significant custom development.

  • Real-time Monitoring

    You need to show progress to users, but building real-time monitoring with polling, WebSocket connections, or custom streaming solutions takes weeks.

Why aipartnerupflow?¶

  • Unified Interface

    One framework handles traditional tasks, HTTP/REST APIs, SSH commands, Docker containers, gRPC services, WebSocket communication, MCP tools, and AI agents—all through the same ExecutableTask interface.

  • Start Simple, Scale Up

    Begin with a lightweight, dependency-free core. Add AI capabilities, A2A server, CLI tools, or PostgreSQL storage only when you need them. No forced upfront installations.

  • Language-Agnostic Protocol

    Built on the AI Partner Up Flow Protocol, ensuring interoperability across Python, Go, Rust, JavaScript, and more. Different language implementations work together seamlessly.

  • Production-Ready

    Built-in storage (DuckDB or PostgreSQL), real-time streaming, automatic retries, state persistence, and comprehensive monitoring—all included. No need to build these from scratch.

  • Extensive Executor Ecosystem

    Choose from HTTP/REST APIs (with authentication), SSH remote execution, Docker containers, gRPC services, WebSocket communication, MCP integration, and LLM-based task tree generation.

What Happens When You Use aipartnerupflow?¶

Before After
Weeks of custom coordination code Days to define task trees with dependencies
Multiple orchestration systems for different execution methods One unified interface for all execution methods
All-or-nothing decisions requiring complete rewrites Gradual addition of AI agents incrementally
Weeks building custom polling or streaming solutions Built-in real-time streaming via A2A Protocol
Manual recovery logic and lost progress Automatic retries with exponential backoff and state persistence
Worrying about resource usage at scale Production-ready from day one, handle hundreds of concurrent workflows

Quick Start¶

New to aipartnerupflow? Get up and running in minutes!

Installation:

pip install aipartnerupflow

Quick Start Guide

Core Concepts

Examples

Documentation Sections¶

Learning Paths¶

🚀 Quick Start (15 minutes)¶

Perfect for getting started quickly:

  1. Quick Start - Get running in 10 minutes
  2. Basic Examples - Try examples
  3. Core Concepts - Understand basics

📚 Complete Beginner (1-2 hours)¶

Step-by-step learning path:

  1. Getting Started Index - Overview and learning paths
  2. First Steps Tutorial - Complete beginner tutorial
  3. Task Trees Tutorial - Build task trees
  4. Dependencies Tutorial - Master dependencies
  5. Core Concepts - Deep dive
  6. Basic Examples - Practice

💼 Professional Developer (2-4 hours)¶

For experienced developers:

  1. Quick Start - Quick refresher
  2. Task Orchestration - Master orchestration
  3. Custom Tasks - Create executors
  4. Best Practices - Learn patterns
  5. API Reference - Complete reference

🔧 Contributor (4+ hours)¶

For framework contributors:

  1. Development Setup - Set up environment
  2. Architecture Overview - Understand design
  3. Contributing - Learn process
  4. Extending - Extend framework

For Users¶

  • Task Orchestration - Complete guide to task orchestration, dependencies, and priorities
  • Custom Tasks - Guide to creating custom tasks with ExecutableTask interface
  • CLI - Complete CLI usage guide
  • API Server - API server setup and usage guide
  • Best Practices - Best practices and recommendations
  • FAQ - Common questions and troubleshooting

For Developers¶

  • Python API - Core Python library API reference (TaskManager, ExecutableTask, TaskTreeNode, etc.)
  • HTTP API - A2A Protocol Server HTTP API reference
  • Quick Reference - Cheat sheet with common snippets
  • Extending - Guide for extending the framework (custom executors, extensions, hooks)
  • Contributing - Contribution guidelines and process

Architecture & Design¶

Examples & Tutorials¶

Quick Navigation¶

By Task¶

I want to...

By Role¶

I am a...

Additional Resources¶

Need Help?¶

Check out our FAQ for common questions and answers, or start a discussion on GitHub.

Ready to start? → Getting Started → or Quick Start →