Home Archive Issue #584
Issue #584

API Generation, AI Agents and File I/O

Nine .NET reads cover SDK-tested agents, calibrated LLM outputs, messaging options, resumable uploads, JWTs, web search, and C# discards.

Jasen's top three picks

  1. 1
    Beyond CRUD Scaffolding: Generating a Structured .NET API

    Medium

    A look beyond basic CRUD toward generating an API with a more intentional structure.

  2. 2
    I Compiled Five .NET Agent Tutorials Against the Real SDKs: What Worked and What Was Invented

    Medium

    A welcome reality check on agent samples: compile the code against current SDKs before adopting the advice.

  3. 3
    Your LLM Is 99% Sure and 95% Right: Calibrated Confidence in .NET 10

    Medium

    Confidence scores are only useful when calibrated; this applies practical statistical guardrails to LLM routing.

Editor's note

I pulled together another deliberately varied set of .NET reads rather than forcing a theme. Meziantou’s file I/O measurements are useful grounding for async decisions, while the agent tutorial audit matters because SDK examples need to survive real compilation.

Today's articles

Related issues

📬 Save the search. Get the .NET stories worth reading at 7 AM ET.