Slide Examples

This repository contains example implementations using Slide's AI agent frameworks: Tyler, Lye, and Narrator.

Overview

These examples demonstrate how to build AI agents using Slide's ecosystem:

• Tyler - The core agent framework for building conversational AI assistants
• Lye - Tool library providing web search, file operations, image processing, and more
• Narrator - Persistence layer for managing conversation threads and file storage

Examples

1. Basic Agent (agent.py)

A complete example showing:

• Thread-based conversation management
• Tool integration (web, image, and file tools)
• Persistent storage with Narrator
• Weave tracking for observability

python agent.py

2. Streaming Agent (agent_streaming.py)

Demonstrates real-time streaming responses:

• Event-based streaming output
• Token-by-token response handling
• Simplified setup without tools

python agent_streaming.py

3. Wandb Inference Examples

Examples configured to use Weights & Biases infrastructure:

• agent_wandb_inference.py - Basic agent with Wandb API
• agent_streaming_wandb_inference.py - Streaming agent with Wandb API

These examples use your WANDB_API_KEY as the OpenAI API key for seamless integration with W&B's proxy.

4. Tyler Chat CLI Configuration

The tyler-chat-config.yaml file shows how to configure the Tyler CLI chat interface with:

• Custom agent identity and purpose
• Model parameters
• Tool selection and configuration

Installation

This project uses uv for Python package management:

# Clone the repository
git clone https://github.com/adamwdraper/slide-examples.git
cd slide-examples

# Install dependencies with uv
uv sync

Environment Setup

Create a .env file with your API keys:

# For OpenAI
OPENAI_API_KEY=your_openai_api_key

# For Wandb examples (optional)
WANDB_API_KEY=your_wandb_api_key

# For Narrator persistence (optional)
NARRATOR_DATABASE_URL=your_database_url

Project Structure

slide-examples/
├── agent.py                           # Full-featured agent example
├── agent_streaming.py                 # Streaming response example
├── agent_wandb_inference.py           # Wandb-integrated agent
├── agent_streaming_wandb_inference.py # Wandb streaming agent
├── tyler-chat-config.yaml             # Tyler CLI configuration
├── pyproject.toml                     # Project dependencies
└── README.md                          # This file

Key Features Demonstrated

• Conversation Management: Thread-based conversations with message history
• Tool Integration: Web search, file operations, and image processing capabilities
• Streaming Responses: Real-time token streaming for responsive interactions
• Persistence: Store conversations and files with Narrator
• Observability: Weave integration for tracking and debugging
• Flexible Configuration: YAML-based configuration for the Tyler CLI

Dependencies

• Python 3.12+
• slide-tyler ≥ 1.3.0
• slide-lye ≥ 0.3.0
• slide-narrator ≥ 1.0.1
• python-dotenv (for environment variables)

Contributing

Feel free to submit issues or pull requests if you have suggestions for improving these examples or adding new ones.

License

This project is open source and available under the MIT License.

Related Links

• Tyler Documentation
• Lye Tools Documentation
• Narrator Documentation
• Weights & Biases