Bring pets to work day today 😃

Still find myself fixing typos and adding question marks (when clearly a question) in Claude etc prompts!

Managing a series of cascading PRs with Claude is so useful....including monitoring CI, fixing issues found, rebasing the various branches and pushing/retrying again. And all from my phone thanks to /remote-control.

Introducing Claude Opus 4.7 www.anthropic.com/news/claude-...

AI-focused sessions are now live for KotlinConf’26! Join experts exploring scalable agents high-quality AI development, and Kotlin-powered machine learning across platforms. Browse the full schedule and stay ahead of the curve! ➡️ https://kotl.in/conf26-schedu

Trying out Claude Code + local Ollama model for something a bit more involved ("support viewing information from api.climatetrace.org/v7/docs/inde...) ....did a pretty good job! Started with empty Compose Multiplatform project. ollama launch claude --model qwen3.5:35b-a3b-coding-nvfp4

Still trying local LLMs on M5 Max device here. Was comparing following and was interesting to see explanation from Gemini below about why Qwen model is faster (and certainly seeing that's the case here in testing). - gemma4:26b-a4b-it-q4_K_M - qwen3.5:35b-a3b-coding-nvfp4

llmfit (github.com/AlexsJones/l...) is nice way to see what models run best on your hardware

Confetti now has the AndroidMakers data thanks to @zsmb.co 🙏 confetti-app.dev

M5 Max laptop arrived and starting to try out various local models (as part of Koog AI agent using Ollama integration). Latency numbers looking pretty good so far. This one is a 4B parameter model but working really well for example I'm using (e.g. very reliable tool calling)

The world of #AI is expansive and ever changing. Most of it filters down for all practical purposes to a couple of key concepts and tools. It's more than sufficient to know those and expand your knowledge if/when needed.

Interesting Fragmented Podcast episode on Background AI Agents fragmentedpodcast.com/episodes/309/

Another useful Claude prompt! "add comments to all the Kotlin files....write in way to use as "speaker notes" during live coding session"

Very cool to see! Showing this data in Claude Desktop using MCP Server from the ClimateTraceKMP sample (github.com/joreilly/Cli...). That makes use of the Kotlin MCP SDK (github.com/modelcontext...) and Ktor to connect to the @climatetrace.org endpoint.

Ollama 0.20.0-rc0 available and with it the new Google DeepMind Gemma 4 local model. Trying it here as part of a Koog AI agent (screenshot thanks to Koog's OpenTelemetry langfuse integration)

And then there were 14! (from github.com/joreilly/Peo... #KMP/#CMP sample)

This is either brilliant or scary: Anthropic accidentally leaked the TS source code of Claude Code (which is closed source). Repos sharing the source are taken down with DMCA. BUT this repo rewrote the code using Python, and so it violates no copyright & cannot be taken down!

Tried new qwen3.5:4b-nvfp4 Ollama model on M1 Max here (in project where it's used with Koog AI agent)...38% faster than qwen3.5:4b (averaged over 5 runs of the agent). I asked Gemini and it initially claimed that I shouldn't see increase in performance like this...asked again and got following 😀

The latest 2.4.0 dev build of Swift Export now supports exposing StateFlow directly to Swift code. Chip-8 sample updated to use it. #KMP github.com/joreilly/chi...

KotlinConf 2026 is packed with inspiring talks in multiple tracks! Not sure how to plan your agenda? Our new blog post highlights a few sessions you might want to check out. 👀 https://blog.jetbrains.com/kotlin/2026/03/kotlinconf-2026-talks-schedule/

Compose Hot..Swan is finally here! 🦢 Compose HotSwan enables Compose Hot Reload in real Android devices, instant UI updates on your running Android app, with state preserved. No rebuilds, no relaunching, no navigating back to the same screen again. hotswan.dev

Loving the Claude/Slack integration.....so nice to be able to use for querying for example history of discussion on particular topic or conclusions reached etc

Good news from kotlinx-rpc! They now have a dev preview available where you can use gRPC with Kotlin Multiplatform! Check out the release notes and resources to try it:

The KotlinConf’26 schedule is live. ✨ Discover upcoming talks on Kotlin’s evolution, AI-powered development, and real-world data workflows. Join us to learn, connect, and grow! ➡️ https://kotl.in/conf26-schedu

The interactive chart support in Claude is really nice. This is based on data from an MCP server in the ClimateTrace KMP sample (github.com/joreilly/Cli...)

Increase Productivity with Git Worktrees as an Android Developer by Domen Lanišnik medium.com/@domen.lanis...

New Room 3.0 alpha release now supports Kotlin/Wasm #KMP developer.android.com/jetpack/andr...

Great to see that support for the OpenTelemetry feature for functional agents is working now with the latest version of Koog. This is example here using LangFuse (cloud.langfuse.com).

Fault tolerant AI Agents on the JVM with Koog framework by @vadim-briliantov.bsky.social www.youtube.com/watch?v=2l1G...

KMP devs 👀 We just opened the beta for Kotzilla KMP Crash Reporting. Built KMP-first: • Unified Android + iOS crash visibility • Readable shared-module stack traces • Lightweight SDK You can try it: kotzilla.io Reply, DM me, or just jump in. #Kotlin #KMP #AndroidDev

Schedule for @kotlinconf.com 2026 now available in Confetti ( github.com/joreilly/Con...)

The @climatetrace.org APIs had some big enough changes between v6 and v7 but migrations like this so much easier now with likes of Claude Code (using OpenAPI spec changes)! "update ClimateApiClient to use v7 of climatetrace apis (api.climatetrace.org/v7/docs/inde... github.com/joreilly/Cli...

A major requirement buying laptops these days will be how well they run local LLMs

Remote control now available for Claude Pro accounts!

All the Kotlin versions in a single blog post! (or almost all ^^) mbonnin.net/2026-02-22-k...

"157 models. 30 providers. One command to find what runs on your hardware" by @axjns.dev github.com/AlexsJones/l...

Have main outline/"scaffolding" in place now for my Live Coding talk at @kotlinconf.com this year....really happy with how it's turning out.

Updated first of #KMP samples I have to use Swift export (given Flow support now available in recent Kotlin 2.4 dev builds). Key usage here is that data is exposed from shared code as a Kotlin StateFlow and observed in iOS app using Swift's for await. github.com/joreilly/chi...

📣 The full #KotlinConf26 schedule is now live on our website! Explore the talks, workshops, and speakers, and start planning your agenda today. Check it out: https://kotl.in/conf26-schedu

Your builds and remixes have brought us to the finish line. 🏁 Because of your collective effort, the #GoogleIO dates are unlocked. Join us May 19-20, 2026. 🗓️ Register today → goo.gle/io2026reg-6

I started programming as a hobby in 1982 (using BASIC on a ZX81) and then professionally from 1990. Industry has of course seen several major transitions since then but probably none as big as what we're experiencing right now. Will be very interesting to see what future holds.

Best Local LLMs to Run On Every Apple Silicon Mac in 2026 apxml.com/posts/best-l...

Roses are red, Violets are blue, I put my heart in commonMain, So it runs where you do. 😀 #KMP

Anyone aware of any comparisons of Claude Code and Junie....with both using same LLM (e.g. Opus 4.6)?

Updated Koog agent in ClimateTrace KMP sample to use a sub agent (that wraps tool to get population data). Seems like this approach provides option to 1. Reduce context 2. Potentially use different LLM (maybe sub agent can work with simpler/cheaper one) github.com/joreilly/Cli...

Looks like the Junie AI coding agent now provides option to use Claude Opus 4.6

Instant Compose - "Start building your Compose app within a minute using any editor, share it on GitHub minutes later." emilflach.github.io/instant-comp...

I asked Junie to make use of more of the @climatetrace.org APIs available (at api.climatetrace.org/v6/swagger/i...) in the ClimateTraceKMP sample (github.com/joreilly/Cli...) and it added UI/requests etc to allow showing more detailed "asset" information....very interesting to see!

I asked Gemini about best way to to manage pricing of FormAI app (play.google.com/store/apps/d...) to ensure LLM token costs are covered. Nothing too unexpected but lots of good advice. Also recommended I should use @revenuecat.com ....probably time I tried that out anyway!