Kenny Pflug

Engineer Manager at TELIS/GWVS. AI, Cloud-Native and .NET, Speaker, Hobby Shouter and Guitarist @not_brawn

[Kenny Pflug @feO2x]

Forget Vibe Coding, do Guided Coding!

1) Plan iteratively with the Coding Agent
2) Let the Coding Agent implement the plan
3) Review and iteratively guide the Coding Agent towards better design, security, performance, etc.

#AI #CodingAgents #GuidedCoding #VibeCoding

[Kenny Pflug @feO2x]

In German, we say "Geht runter wie Butter":

Thanks to #ADCkonf 2026! I hope to join you again in 2027!

[Kenny Pflug @feO2x]

Light.PortableResults 0.6.0 is a hobby project of mine that I started in January 2026. With the help of #AI #CodingAgents an my own approach called #GuidedCoding, I managed to create this awesome library.

https://t.co/JD1Js5sDrz

[Kenny Pflug @feO2x]

🎇 Light.PortableResults 0.6.0 is out!

This release brings a new Roslyn Source Generator that anaylzes your validators and creates #OpenAPI schema definitions and examples out of it. Let me show you how this works!

https://t.co/qapa1QfqYK

#csharp #dotnet #aspnetcore

[Kenny Pflug @feO2x]

Step 3: benefit from auto-generated #OpenAPI schemas and examples. In the past, I often only properly documented the happy path in OpenAPI - but now, Light.PortableResults takes care of that!

[Kenny Pflug @feO2x]

@lcamtuf @IceSolst I agree. The thing is: who is writing your guardrails? I would guess it's the agent, too? Then you have to review the guardrails - otherwise you can't trust them.

[Kenny Pflug @feO2x]

✨ But things will get even easier: my next task is a Roslyn Source Generator that will automatically create OpenAPI schemas and corresponding examples by inspecting your Validator<T> subclasses. Stay tuned!

https://t.co/BLKABs7HSh

[Kenny Pflug @feO2x]

🦾 Light.PortableResults 0.5.0 is out now!

🚀 Version 0.5.0 brings OpenAPI support to Light.PortableResults. Its now easier to create OpenAPI schemas from Light.PortableResults.Validation assertions. Of course, all this is #dotnet #NativeAOT compatible!
https://t.co/qapa1QfqYK

[Kenny Pflug @feO2x]

🤖 Version 0.5.0 was of course implemented with #GuidedCoding! Across 7 plans, I implemented OpenAPI integration - more than 8,000 lines of code in less than two weeks in my spare time.

https://t.co/JD1Js5sDrz

[.NET @dotnet]

.NET 10.0.7 OOB dropped to fix a DataProtection regression + CVE.
If you’re on 10.0.0–10.0.6, update ASAP and redeploy.
More info: https://t.co/Q6K9lFRNgG

[Andrew Lock "Sock" @andrewlocknet]

Blogged: Running AI agents with customized templates using docker sandbox

https://t.co/bFZBrfvtWp

In this post I describe how to create custom templates for Docker Sandboxes, so that your sandboxes start with additional tools immediately available (like .NET)

#docker #dotnet

[Kenny Pflug @feO2x]

@davidfowl Depth. Building high-performance #dotnet stuff the last few months - has been quite the journey.

https://t.co/qapa1QfqYK

[Kenny Pflug @feO2x]

Lear more about Guided Coding on my website: https://t.co/JD1Js5sDrz

#AI #CodingAgents #GuidedCoding

[Kenny Pflug @feO2x]

🥳 Light.PortableResults 0.2.0 is out now!

🆕 0.2.0 introduces Light.PortableResults.Validation which simplifies error creation. Compared to FluentValidation, it is at least 5 times faster while allocating only 9% of the memory.

#csharp #dotnet
https://t.co/qapa1QfqYK

[Kenny Pflug @feO2x]

🤖 The library was built AI-first. I used an approach I call "Guided Coding" to write more than 90% of the code base with AI agents while staying in full control. The result: ~41,000 lines of C# code, ~102,000 total lines in the repo since January 5th 2026 - as a side project.

[Kenny Pflug @feO2x]

If this project evolves a little bit more, this can save so much money and allow for more interactions within your 5-hour/daily/weekly limits.

[Kenny Pflug @feO2x]

Check out rtk! It's a proxy for Coding Agent CLI calls and reduces the amount of tokens passed to the LLM by 60% to 80%. Get more value out of your subscription.

#AI #CodingAgents

https://t.co/TBe5NklqCD

[Kenny Pflug @feO2x]

Tried out @JetBrainsRider 2026.1 yesterday and this morning. Good job, especially to the performance team! It feels way snapier and the #AI Integration with options for Claude Agent and Codex is getting more and more awesome!