Playbooks AI

Finally, Natural Language Programming is here!

Playbooks AI™ lets you program AI agents using plain English instead of code. Our patent-pending engine turns human-readable instructions into executable AI behavior — no coding required.

Status: Playbooks AI is still in early development. We're working hard and would love your feedback and contributions.

Table of Contents

• Show me!
• What is Natural Language Programming?
• Features
• How it works
• Quick start
• Who Should Use Playbooks?
• Roadmap
• Contributing
• License
• Contributors

Show me!

Here's the simplest Natural Language Program you can write — a Hello World agent:

# Hello World!

## Say Hello

### Trigger
At the beginning

### Steps
- Greet the user with a friendly "Hello, World!" message.

That's it! Just plain English that both humans and AI can understand. Here "Hello World!" is name of the AI agent, while "Say Hello" is a playbook. A playbook is executed as soon as its trigger condition is satisfied.

Now, let's look at something more powerful. Here's a web search agent that:

• Uses search engines when needed
• Performs deep research to gather information
• Filters out inappropriate topics

See Playbooks AI implementation of a Web Search Chat agent — about 50 lines of English instructions.

Now compare that with an equivalent LangGraph implementation — about 200 lines of complex code that's harder to understand and modify.

🔗 Ready to write your first Natural Language Program? Get started here.

What is Natural Language Programming?

Natural Language Programming lets you create AI applications by writing instructions in plain English. Think of it as pseudocode that actually runs.

Building AI agents today forces you to choose between three frustrating options:

1. Writing complex code → Requires technical expertise
2. Using no-code UI builders → Gets messy for complex workflows
3. Direct prompting → Results in unpredictable behavior

With Playbooks AI, you simply write clear instructions in a playbook format that:

• Business people can read and modify
• AI can execute reliably
• Handles complex logic, tool usage, and multi-agent collaboration

Natural Language Programming vs. Traditional Approaches

Feature	Playbooks AI 🏆	Code-Based Frameworks	UI-Based Agent builders	Direct Prompting
Ease of Use	✅ Write in plain English	❌ Requires Python expertise	✅ No-code UI, but gets messy	✅ Just type a prompt
Behavior Control	✅ Easily modify agent behavior	❌ Requires coding to change	❌ Hard to translate requirements into UI	❌ Unpredictable results
Workflow Complexity	✅ Handles simple & complex logic	✅ Handles complex logic, but requires coding	❌ Hard to scale beyond simple workflows	❌ No structured execution
External API Calls	✅ Simple tool calling	✅ Explicit tool calling	✅ Often requires prebuilt integrations	❌ Manual copy-pasting, no automation
Scalability	✅ Designed for 100s-1000s of playbooks	✅ No limit, but code complexity grows	❌ UI becomes unmanageable at scale	❌ Cannot scale beyond one-off conversations
Business User Friendly	✅ Yes	❌ No, requires coding	❌ No, complex workflow graphs	❌ No, requires prompt engineering

Features

Write programs in plain English

• Define AI agent behavior using natural language instead of code
• Let non-technical team members understand and modify agent behavior
• Talk with a copilot to improve your natural language programs

Powerful execution engine

• Playbooks AI faithfully follows your instructions
• Build complex behavior using hundreds or thousands of playbooks
• Easily create multi-agent systems
• Call external tools with simple language
• Dynamic triggering to handle special cases and validations
• Respond to external events

Build any AI application

• Create a wide range of applications:
  • Intelligent chatbots
  • Customer support agents
  • Virtual assistants
  • Team automation tools
  • Workflow automation

What will you build with Playbooks AI?

Quick start

2 easy ways to try Playbooks AI:

1. Visit runplaybooks.ai and try out the demo playground, OR

2. On command line

pip install playbooks
poetry run python src/playbooks/applications/agent_chat.py examples/playbooks/chat.md --stream

Who Should Use Playbooks?

Natural Language Programming with Playbooks AI is perfect for:

✅ Developers & Engineers – Create AI agents without writing complex state machines

✅ Business Teams – Modify AI behavior without coding or technical expertise

✅ Product Managers – Quickly prototype and iterate on AI features

✅ AI Researchers – Experiment with multi-agent systems more efficiently

✅ Automation Specialists – Build intelligent workflows with API integrations

Roadmap

We're just getting started! Here's what's coming next:

• Playbooks Observer for observability and debugging
• Online planning by generating playbooks
• Process multiple trigger matches simultaneously
• Playbooks Hub for community sharing
• VSCode extension for debugging
• Copilot for conversational playbook creation
• Multi-agent communication
• Inference speed optimizations
• Tool sandboxing
• PlaybooksLM fine-tuned model
• Playbooks Platform with enterprise features

Contributing

Welcome to the Playbooks community! We're excited to have you contribute.

If you want to help, checkout some of the issues marked as good-first-issue or help-wanted found here. They could be anything from code improvements, a guest blog post, or a new cookbook.

Development Environment Setup

1. Clone the Repository

   git clone https://github.com/playbooks-ai/playbooks.git
   cd playbooks

2. Environment Variables Set up environment variables for the playbooks package (.env):

   cp .env.example .env

   Edit .env to configure LLM and API settings.

3. playbooks Python package Setup

   # Create and activate a virtual environment (recommended)
   
   python -m venv venv # or conda create -n venv python, or pyenv virtualenv venv
   
   source venv/bin/activate  # On Windows: venv\Scripts\activate
   
   # Install playbooks Python package in development mode
   pip install poetry
   poetry install

Testing

We use pytest for testing. Here's how to run the tests:

1. Run playbooks Python Package Tests

   pytest

Getting Help

• Join our Discord community
• Check existing issues and discussions
• Reach out to maintainers

We appreciate your contributions to making Playbooks better! If you have any questions, don't hesitate to ask.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Contributors

This project is maintained by Playbooks AI.