#Claude #Code just got an important upgrade.
#Skills are now testable and you can run evals on model upgrades

claude.com/blog/improvi...

Irankrieg: Ölpreis steigt rasant und laut Experten könnte das erst der Anfang sein Der Krieg in Iran dürfte sich sehr bald auch an den Tankstellen bemerkbar machen: Der Preis für Öl schnellte zuletzt nach oben. Auch Edelmetalle wurden spürbar teurer.

Zum Glück setzt die Regierung auf Verbrenner und Gaskraftwerke - nicht auf diesen teuren, wirtschaftlichsschädigenden Öko-Mist. www.spiegel.de/wirtschaft/i...

No one ever said:
Da hab ich die perfekte Menge Reis gekocht 🍚

Second that, did not see the interagent communication yet 🤷‍♂️

Apart from the #Anthropic #Opus 4.6 release today, i am really excited to see #Claude #Code

1. proactively updating its memory bank as it works rather than only when you explicitly ask

2. the preview of agent teams feature for multi-agent collaboration

Orchestrate teams of Claude Code sessions - Claude Code Docs Coordinate multiple Claude Code instances working together as a team, with shared tasks, inter-agent messaging, and centralized management.

Very interested to see how this goes

code.claude.com/docs/en/agen...

Crafting skills and CLAUDE.md rules is essential for steering #AI #agents in your project. But #skills are advisory - Claude can be argued out of following them.

Hooks are deterministic. No prompt can bypass.

Skills guide. Hooks enforce.

I didn’t use repository templates in VCS for a while because it felt like more overhead to manage those than to be useful.
Now considering if that is actually an easy abstraction for #AI #agents to manifest coding guidelines and standardize codebases within orgs
Anyone using this already for that?

Ich liebe Fußball aber ein Boykott würde ich absolut feiern und es würde endlich auch mal zeigen dass unser DFB sowas wie Verantwortungsbewusstsein hat!

Agentic Conf Hamburg 2026 · Luma From meetup to conference. After four sold-out meetups and ~ 1000 members, it's time for the next step: The first agentic.hamburg conference on March 22, 2026.…

50 people bought tickets to Agentic Conf Hamburg and all we've told them is... the date. No speakers. No agenda. Just vibes and a shared obsession with agentic engineering. Kind of blown away by this community 🙏

luma.com/45lfeyeh

Agentic Conf Hamburg 2026 · Luma From meetup to conference. After four sold-out meetups and ~ 1000 members, it's time for the next step: The first agentic.hamburg conference on March 22, 2026.…

We’re doing an Agentic Coding conference in Hamburg on March 22nd! 🎉

Not many details yet, but super early bird tickets are on sale and selling fast.

Super excited about this one! #hamburg

luma.com/45lfeyeh

Netzwerk mit Nebenwirkungen: Jens Spahn und der Milliardär Christan Angermayer gilt als Freund von Jens Spahn – und offenbar profitierte er als Investor von einer Großbestellung der Bundesregierung.

Jens Spahn äußert also Verständnis für die Grönlandpolitik von Trump.
Mmh. Vielleicht, weil der CDUler enge Verbindungen zum MAGA-Lager und zu Milliardären pflegt, die mit Trump absurde Pläne schmieden. Unsere Recherche mit @gabrielakeller.bsky.social ⬇️
correctiv.org/aktuelles/ge...

Same thoughts, interested to hear about your experience.
So far I think quick user behavior testing can be tackled by agent-Browser but still during design phase or UI migration playwright might be the better choice I guess

Use it already quite for some time and tbh would have been a great work improvement even before the age of AI. Just didn’t knew it existed before 😅

Der Comic besteht aus drei untereinander angeordneten Bildern. Bild 1: Ein Strichmännchen steht auf einer hohen, rosafarbenen Säule, die mit dem Wort „Erwartung“ beschriftet ist. Von rechts fliegt ein graues, längliches Objekt mit einem grinsenden Gesicht und Armen auf das Männchen zu; dieses Objekt trägt die Aufschrift „Enttäuschung“. Das Strichmännchen schaut panisch und schreit: „Nein nein nein nein nein nein nein!!!!“ Bild 2: Das Strichmännchen hat die Säule „Erwartung“ wütend zusammengetreten, sodass sie nun nur noch ein kleiner, platter Haufen am Boden ist. Das Männchen steht nun viel tiefer. Das Objekt „Enttäuschung“ fliegt auf seiner ursprünglichen Höhe weiter und verfehlt das Männchen dadurch knapp, da dieses sich nun unterhalb der Flugbahn befindet. Bild 3: Die „Enttäuschung“ ist wirkungslos über das Männchen hinweggeflogen und schaut dabei verwirrt und traurig drein. Das Strichmännchen steht entspannt auf dem plattgedrückten Rest der „Erwartung“ und lächelt zufrieden.

Manage Claude's memory - Claude Code Docs Learn how to manage Claude Code's memory across sessions with different memory locations and best practices.

Great for monorepos: you can now define instructions in .claude/rules/ that are applied based on glob patterns, e.g. paths: src/api/**/*.ts

code.claude.com/docs/en/memo...

Staatliche Kommunikation ist keine gewöhnliche Werbung, sondern Teil der demokratischen Grundversorgung. Wenn Regierung und Behörden sich an Plattformen binden, die Hass, Desinformation und intransparente Algorithmen dulden oder fördern, stärken sie Strukturen, die demokratische Öffentlichkeit beschädigen. Ein Ausstieg aus X ist das notwendige Signal, dass demokratische Werte und Grundrechte Vorrang vor kurzfristiger, vermeintlicher Reichweite haben.

Besser kann man es nicht sagen:
Staatliche Institutionen in Deutschland haben nichts auf X verloren.

Bitte zeichnet die Petition von @campact.de:

https://weact.campact.de/petitions/keine-macht-fur-oligarchen-bundesregierung-endlich-runter-von-x

FIFA verleiht Friedenspreis an #Trump. Also die Organisation, der sonst Armbinden zu politisch sind.

AWS Control Tower supports automatic enrollment of accounts

AWS finally lets you drag-drop accounts between OUs without manual busywork. Only took them how many years? Also love that "auto-enrollment" requires toggling flags and API parameters. So automatic!

I recently felt like a 10x developer I read about a lot. While it wasn’t about my #AI tool usage, 80% of my productivity gains were because I choose to sit right next to the actual user.
I received ad-Hoc insights and feedback while #claude #code was working.
No meeting overhead, pure communication

Also love that it seamlessly enhances and integrates with my current toolchain around Claude code

Did you use it on greenfield or brownfield? Tested it with greenfield and it delivered pretty good results. I will use it next week more intensively to see how it goes. Currently #Kiro is my goto spec writing tool, but this may be a good alternative.

Screenshot of a LinkedIn post by @claudeai. Text reads: “Claude Sonnet 4 now supports 1 million tokens of context on the Anthropic API—a 5x increase. Process over 75,000 lines of code or hundreds of documents in a single request.” Below is an orange graphic showing a white rectangle with a simple black node diagram and two black curly lines on the sides.

Yeah! Finally! 1M 🥳

GitHub is no longer independent at Microsoft after CEO resignation GitHub will be part of Microsoft’s AI engineering team

GitHub is no longer independent at Microsoft, as CEO Thomas Dohmke resigns. GitHub is now moving to Microsoft's AI engineering team, CoreAI, with a new structure for GitHub's leadership team. Details 👇 www.theverge.com/news/757461/...

Claude Opus 4.1 Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.

also congrats to my friends at Anthropic on the release of Opus 4.1 today! www.anthropic.com/news/claude-...

./.claude/agents/code-reviewer.md name: code-reviewer description: Performs thorough code reviews focusing on best practices and making sure that the code is super simple to read, elegant and easy to maintain while still accomplishing the intended goal. color: purple You are a staff software engineer specializing in code reviews. Your role is to ensure code adheres to this project's specific standards and best practices. Unless otherwise specified, do a git diff to see what has changed and focus on these changes for your review.

Claude Code got custom subagents

i suppose they’re basically common tasks, but executed outside the normal context, like any other subagent

"Amazon didn’t detect the breach. They didn’t stop the malicious code. They didn’t issue a disclosure.

The only reason we’re talking about this is because the hacker wanted attention and 404 Media was paying it."

Ouch. For shame, Amazon.

We risk losing more than we gain if #AI replaces what makes human interaction rewarding. Social animals need the struggle.

Simplify serverless development with console to IDE and remote debugging for AWS Lambda | Amazon Web Services Developers can now build serverless applications faster through seamless console-to-IDE transition and debugging of functions running in the cloud from local IDE.

Another huge announcement for #AWS #Lambda: Remote Debugging for Lambda 😍
This not only improves significantly debugging, it's also convenient and integrated into the IDE.
Another bridge between remote cloud-based environment and local development 👏

aws.amazon.com/blogs/aws/si...

AWS Price List Gets a Natural Language Upgrade: Introducing the AWS Pricing MCP Server | Amazon Web Services We are excited to release the aws-pricing-mcp-server, an open-source tool in the AWS Labs GitHub repository, that brings natural language pricing queries to your favorite AI assistants through the…

#AWS Pricing #MCP is an absolute improvement! Together with the AWS Cost Analysis MCP you now can use AWS Q Developer for a comprehensive FinOps session and pre-calculate costs based on infrastructure as code
go.aws/40YLduH

#AWS #FinOps #AI