PromptifyNET 1.0.0

PromptifyNET

A lightweight, config-first prompt chaining library for .NET that makes LLM-powered application development simple, fast, and flexible.

Features

• Config-first prompt chaining and orchestration
• Multiple LLM provider support (OpenAI, Anthropic, OpenRouter, and more)
• Built-in prompt templates and chaining logic
• Extensible plugin system for custom steps
• Minimal code required for setup and usage
• Async/await and DI-friendly
• Logging and diagnostics support
• High performance and thread safety

Installation

dotnet add package PromptifyNET

Quick Start

using PromptifyNET;

// Create the engine
var engine = new PromptChainEngine()
    .RegisterProvider(new DummyLlmProvider());

// Configure your prompt chain
var config = new PromptChainConfig
{
    Steps = new List<PromptStep>
    {
        new PromptStep
        {
            Name = "greeting",
            PromptTemplate = "Hello, {{name}}!",
            Provider = "Dummy"
        }
    }
};

// Run the chain
var context = new Dictionary<string, object>
{
    ["name"] = "World"
};

var result = await engine.RunAsync(config, context);
Console.WriteLine(result); // Output: [LLM Response to: Hello, World!]

Advanced Usage

Multi-step Chains

var config = new PromptChainConfig
{
    Steps = new List<PromptStep>
    {
        new PromptStep
        {
            Name = "generate",
            PromptTemplate = "Generate a number: {{input}}",
            Provider = "Dummy"
        },
        new PromptStep
        {
            Name = "double",
            PromptTemplate = "Double the number: {{last}}",
            Provider = "Dummy"
        }
    }
};

Using Plugins

var engine = new PromptChainEngine()
    .RegisterProvider(new DummyLlmProvider())
    .RegisterPlugin(new UppercasePlugin());

var config = new PromptChainConfig
{
    Steps = new List<PromptStep>
    {
        new PromptStep
        {
            Name = "greeting",
            PromptTemplate = "Hello, {{name}}!",
            Provider = "Dummy",
            Plugins = new List<string> { "Uppercase" }
        }
    }
};

Available Providers

• DummyLlmProvider: For testing and development
• OpenAiProvider: OpenAI API integration
• AnthropicProvider: Anthropic Claude API integration
• OpenRouterProvider: OpenRouter API integration

Available Plugins

• UppercasePlugin: Converts text to uppercase
• LowercasePlugin: Converts text to lowercase
• SummarizePlugin: Creates text summaries
• TranslatePlugin: Translation placeholder
• SentimentPlugin: Basic sentiment analysis
• FormatPlugin: Formats text (Markdown, HTML, JSON)
• FilterPlugin: Filters text by length

Creating Custom Providers

public class CustomProvider : ILlmProvider
{
    public string Name => "Custom";
    
    public async Task<string> InvokeAsync(string prompt, IDictionary<string, object>? context = null)
    {
        // Your custom LLM integration logic here
        return $"Custom response to: {prompt}";
    }
}

Creating Custom Plugins

public class CustomPlugin : IPromptPlugin
{
    public string Name => "Custom";
    
    public async Task<string> ExecuteAsync(string input, IDictionary<string, object>? context = null)
    {
        // Your custom processing logic here
        return $"Processed: {input}";
    }
}

Configuration

PromptifyNET supports various configuration options through the context parameter:

var context = new Dictionary<string, object>
{
    ["targetLanguage"] = "Spanish",
    ["format"] = "markdown",
    ["maxLength"] = 500
};

Contributing

We welcome contributions! Please feel free to submit issues and pull requests.

License

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