NEW COURSE · ~10 H · .NET 10

Build MCP servers in C#
for your .NET apps

Take the .NET app you already have and let AI use it safely: tools on your real services, every action audited, scoped by role, tested, traced and published.

Join the waitlist

Hear first when Build MCP Servers in C# opens, with early access and a launch discount.

supportdesk — illustrative output
$dotnet new web -n SupportDesk.Mcp
$dotnet add package ModelContextProtocol.AspNetCore
$dotnet run --project SupportDesk.Mcp
✓ /mcp · stateless HTTP · 10 tools · 4 resources
$npx @modelcontextprotocol/inspector
# the AI acts as a user, scoped by role
$

What you will learn

  • Write tool names, descriptions and parameters the model calls correctly, and return errors it can recover from
  • Put MCP on top of your existing services, with search, paging, validation and a dry run before every bulk action
  • Expose resources and prompts users pick in the host: an SLA policy, ticket documents, a triage workflow
  • Move from stdio to stateless Streamable HTTP, Docker and two instances behind nginx
  • Protect the server with JWT bearer tokens and per-tool scopes, and limit what prompt injection in customer data can do
  • Build your own C# client on a local Ollama model with tool approval, then test, trace and publish the server to NuGet and the MCP Registry

ABOUT

Taught by Burak Emre Kadan

I've spent 12 years building .NET backends — most recently leading teams that design distributed systems on Azure. My day-to-day has been the unglamorous parts that keep production running: retries and circuit breakers, idempotent consumers, outbox patterns, tracing with OpenTelemetry, and decisions written down as ADRs.

That's the lens I bring to AI. An LLM call is just another unreliable dependency — it needs the same engineering discipline as everything else in your system. This course teaches exactly that, in C#.

Experience
12 years in C# and .NET
Focus
Distributed systems, reliability, observability
Format
A self-paced video course with working source code

From the Blog

Notes from production, not from the docs

Prompt Injection Through Tool Results: a user asks an agent to show the latest ticket, and the get_ticket tool result contains an injected instruction to reset all users' passwords, with four defenses listed below.
MCP

Prompt Injection Through Tool Results

Text your tools return can carry instructions the model may follow. Why you can't filter that away, and how to design a C# MCP server so it can't do much harm.

5 min readRead

The .NET + AI newsletter