Why should you care?

Tool-using agents need a clear kernel, not a pile of ad-hoc HTTP calls.

See it live — copy this example

Run CLI steps in a terminal. Run C# samples in a .NET 8 console or Web API project with Ollama running on localhost:11434.

// Planner-Style Tool Loops — Semantic Kernel style sketch
var builder = Kernel.CreateBuilder();
builder.AddOllamaChatCompletion("phi3:mini", new Uri("http://localhost:11434"));
var kernel = builder.Build();
// Register plugins/tools here for Planner-Style Tool Loops
var answer = await kernel.InvokePromptAsync(
    "You are LocalAIDesk. Be concise. Task: describe Planner-Style Tool Loops in 2 bullets.");

What happened?

• Kernel targets Ollama.
• InvokePromptAsync runs a constrained prompt.
• Plugins (next lessons) add tools.

Practice next

1. Read the example and rewrite it in your own repo.
2. Run it against local Ollama (or finish the Ollama module first).
3. Change one prompt constraint and compare output.
4. Tighten the prompt to 3 bullets max.
5. Log duration of the model call.

Remember

You can explain Planner-Style Tool Loops simply. You have a runnable local sketch. You know one risk to watch.

Planner-Style Tool Loops in LocalAIDesk

Your team applies planner-style tool loops while building a private agent desk.

Outcome: A concrete next step exists in code or checklist form.