A free ebook for .NET developers who want to build real AI-powered applications — without the hype, without the Python tax, and without spending a fortune on API credits.
You can also download this ebook free from Leanpub.
- 🧠 More C# samples for current AI models: AI models for C# developers
- ✍️ New posts on AI for .NET developers: taswar.zeytinsoft.com
Build real AI features into your .NET apps — in C#, no Python required.
⭐ Star this repo · 📥 Download free on Leanpub · 🐛 Report an issue
git clone https://github.com/taswar/PromptEngineeringForDotNetDevelopers
cd PromptEngineeringForDotNetDevelopers/chapter-04/src/PromptBuilder
dotnet run
- .NET 10 or later
- An IDE (Visual Studio 2026, VS Code with C# Dev Kit, or Rider)
- Either: a free LM Studio install or an OpenAI / Azure API key
- A healthy scepticism of AI marketing copy
Each chapter is a standalone Markdown file with embedded code snippets.
Each src/ folder contains a runnable .NET project for that chapter's practical exercise.
Each tests/ folder contains unit tests for the chapter's code patterns.
Clone the repo, open a chapter, follow along.
No Jupyter notebooks. No Python. Just C#.
git clone https://github.com/taswar/PromptEngineeringForDotNetDevelopers
cd PromptEngineeringForDotNetDevelopers| # | Title | Status | Topics |
|---|---|---|---|
| 1 | The .NET Developer's AI Landscape | ✅ Complete | LLMs, cost spectrum, Microsoft AI stack |
| 2 | Setting Up Your AI Dev Environment | ✅ Complete | LM Studio, GGUF/quantisation, MEAI, OpenAI, Azure AI Foundry, IConfiguration secrets |
| 3 | How LLMs Work (Just Enough Theory) | ✅ Complete | Tokens, context windows, temperature, model families, IChatClient |
| 4 | Anatomy of a Great Prompt | ✅ Complete | 5-part prompt anatomy, two key principles, PromptBuilder fluent class |
| 5 | Core Prompting Techniques | ✅ Complete | Zero-shot, few-shot, CoT, self-consistency, sycophancy, rubric-based prompting |
| 6 | Structured Outputs and Advanced Patterns | ✅ Complete | JSON structured output, defensive parsing, streaming, resilience, model-as-validator |
| 7 | Prompt Patterns for Real Developer Workflows | ✅ Complete | Code review, test generation, commit messages, documentation, DevToolkit app |
The fastest way to see PromptBuilder in action:
cd chapter-04/src/PromptBuilder
# Set your LM Studio model name (get it from GET http://localhost:1234/v1/models)
# Then run:
dotnet runOption A (LM Studio) is active by default. To use OpenAI or Azure AI Foundry, comment out Option A and uncomment your chosen block — instructions are in Program.cs.
# Chapter 2 — unit tests (no LM Studio required)
dotnet test chapter-02/tests/HelloAI.Tests --filter "Category!=Integration"
# Chapter 3 — unit tests (no LM Studio required)
dotnet test chapter-03/tests/ParameterPlayground.Tests --filter "Category!=Integration"
# Chapter 4 — unit tests (31 tests, no LM Studio required)
dotnet test chapter-04/tests/PromptBuilder.TestsThis book is free and will stay free. It is written outside my day job — evenings, weekends, and the occasional early morning with too much coffee.
If it has helped you land a job, ship a feature, or save a debug session at 2 AM — a coffee is a genuinely kind way to say thanks. It keeps me caffeinated and writing.
Not in a position to buy a coffee? No problem. Star the repo, share it with a colleague, or open an issue with a typo you spotted. Any of those helps. 🙂
Prompt Engineering for .NET Developers Copyright © 2026 Taswar Bhatti
This work is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License.
To view a copy of this license, visit: https://creativecommons.org/licenses/by-nc/4.0/