@digitarald

👨‍🎨 Crafting AI/DevX | 🛠 PM on VS @Code | Views are my own, obvs

 Global Citizen in California
digitarald.de
Joined April 2008
  • Tweets4.6K
  • Following1K
  • Followers4.1K
208Media
Harald Kirschner retweeted
Pamela Fox
@pamelafoxSep 16
It was so fun chatting with @digitarald + @TobinSouth at MCP Community Connect on Monday! We talked about the increasing use of MCP by the non-tech crowd, challenges of building MCP registries, MCP for science labs and agriculture, and what success looks like for MCP in 5 yrs.
Pamela with Harald and Tobin

ALT Pamela with Harald and Tobin

1
10
971
Harald Kirschner retweeted
Pamela Fox
@pamelafoxSep 2
We're bringing together an awesome line-up of speakers on September 14th for our free MCP conf at GitHub HQ! Join us to hear from @TobinSouth @digitarald @sofiiiiiasz @ptdamiba @dbreunig @JiquanNgiam @PrathmeshPatel_ @radhigulati @cedricvidal and more: luma.com/global-8rdk
Graphic displaying all the speakers

ALT Graphic displaying all the speakers

2
7
2
21
1,908
Harald Kirschner retweeted
Angie Jones
@techgirl1908Aug 6
Oh we neeeeded this! 🙌🏾 I've been begging for a way to use my Skills and MCPs across the various harnesses I use. Amazon, Cursor, Microsoft, OpenAI, and Vercel got together and created a standard for Agent Plugins. Currently supported in Codex, Cursor, GitHub Copilot, Kiro, and VSCode. Interoperability we love to see! 👏🏾 Hear from the Technical Steering Committee
Agentic AI Foundation
@AgenticAIFdnAug 6
A much needed open standard for agent interoperability is here! 🎉 Agent Plugins gives Agent Skills and MCP servers a shared package format, so developers can build once and use the portable parts across compatible agent clients. The standard was developed by representatives from Amazon, Cursor, Microsoft, OpenAI, and Vercel. Hear from the Agent Plugins Technical Steering Committee about how the standard works and what it means for the agent ecosystem 👇 aaif.io/blog/from-skills-and… (note that the Agent Plugins is independently governed and is not an AAIF project)

Agent Plugins 1.0: Portable Package Format for AI Skills - Agentic AI Foundation (AAIF)

Agent Plugins 1.0 defines an open, vendor-neutral package format for distributing Agent Skills and MCP servers across compatible AI agent clients.

aaif.io
6
11
2
79
16,472
Weekend hobby project energy: give Autopilot the goal, let Fleet orchestrate the long haul, return later to see what your tiny side project has become. GPT-5.5 is 60% off for Copilot Max and 40% off for Pro+ this weekend. xitter.cf/github/status/20781567…
GitHub
@githubJul 17
📣 GPT-5.5 this weekend: 60% off for Copilot Max users and 40% off for Pro+ users. 🎉 Put it toward a tricky bug, a larger change, or updates across multiple files — wherever you use Copilot. No code or opt-in required. Happy coding. ✅
GitHub Copilot app showing GPT-5.5 is 40% off for Pro+ users this weekend, until 00:00 UTC July 20.

ALT GitHub Copilot app showing GPT-5.5 is 40% off for Pro+ users this weekend, until 00:00 UTC July 20.

1
4
1,147
Harald Kirschner retweeted
Burke Holland
@burkehollandJul 11
A public service announcement for containers.dev
4
7
93
11,047
We just published a deep dive on improving token efficiency in GitHub Copilot within VS Code. The headline isn't "use fewer tokens." It's "send better context." code.visualstudio.com/blogs/…

Improving token efficiency for GitHub Copilot in VS Code

Learn how we're improving token efficiency for GitHub Copilot to reduce costs and latency for VS Code users.

code.visualstudio.com
2
6
552
Harald Kirschner retweeted
Oren Melamed
@OrenMeMay 31
🧵 The @GitHubCopilot Token Tax is Here. Are You Ready for UBB? Tomorrow, the rules of the game change. @GitHubCopilot is officially moving from the flat-rate Premium Request Units (PRUs) to Usage-Based Billing (UBB). Here is a list of unfiltered options for you (and also untested ) to analyze your usage and start saving on your precious tokens. A thread 👇
9
11
1
103
17,593
Harald Kirschner retweeted
Pierce Boggan
@piercebogganMay 28
Bring your own key in @code now supports 'custom endpoints' including chat completions, responses, and messages API shapes! code.visualstudio.com/update…

Visual Studio Code 1.122

Learn what's new in Visual Studio Code 1.122

code.visualstudio.com
6
3
2
74
9,518
Practical guide for debugging AI agents in @code with OpenTelemetry. Shows how to go from "agent feels slow" to seeing exact token counts, model latency, and finish reasons per turn. The @aspiredotdev dashboard runs locally in Docker, no accounts needed. opentelemetry.io/blog/2026/g…

Inside the LLM Call: GenAI Observability with OpenTelemetry

Your AI agent just took 45 seconds to answer a simple question. Was it the model? A slow tool call? A retry loop? Every time an application calls an LLM, a chain of model calls, tool invocations, and...

opentelemetry.io
3
1
8
1,575
Claiming my contribution of finding a breaking issue for /remote 10min before I was meant to demo it during VS Code's office hours. Didn't fix it, but issues are a great way to start contributing to OSS. Keep reporting well-written bugs for any issues you see!
Peng Lyu
@njukidrebornMay 18
I'm here to claim my contribution: merging my PM @pierceboggan 's PR to make this happen in @code
1
6
844
Harald Kirschner retweeted
Kenneth Auchenberg 🛠
@auchenbergMay 7
Recently I've had several people telling me that they are moving from @cursor and back to @code. Codex/Claude is doing all of work, and they want a clean code editor to review some of the code in. Interesting times.
20
2
4
48
10,100
Hooks are the most underused AI coding customization. Gate file writes on lint ✓ Inject env context ✓ Block destructive ops ✓ @burkeholland walks through them to make them accessible (and VS Code's customization tools make them debuggable) 👇 youtube.com/watch?v=03CfGf9i…

Completely understand hooks in less than 20 minutes

Hooks let you handle events at key points in the Copilot or Claude ...

youtube.com
5
865
Harald Kirschner retweeted
Milan Jovanović
@mjovanovictechMar 26
GitHub Copilot in Autopilot mode is insanse
25
21
1
407
59,668
Harald Kirschner retweeted
Pierce Boggan
@piercebogganMar 19
Code survival is a great metric for understanding the effectiveness of models on different scenarios in @code! Should we share more metrics like this?
Kyle Daigle
@kdaigleMar 19
Hot take from looking at @github Copilot telemetry: benchmarks make coding models look wildly different. Production workflows make them look much more similar. 👀 We looked at 23M+ Copilot requests and examined one simple metric: code survivability.
10
5
2
73
10,454
Harald Kirschner retweeted
James Montemagno
@JamesMontemagnoMar 8
And my quick port over to add #GitHubCopilot support throught the SDK and the CLI. Had the CLI analyze and add first round of support: github.com/jamesmontemagno/t…
Theo - t3.gg
@theoMar 7
T3 Code is now available for everyone to use. Fully open source. Built on top of the Codex CLI, so you can bring your existing Codex subscription.
6
8
109
12,487
Harald Kirschner retweeted
VS Code Changelog@VSCodeChangelogMar 4
🚀 Insiders Update - Feb 26-Mar 4 14 new features 🧵 ✨ AI CLI terminal profiles now grouped ✨ MCP Apps support file downloads ✨ Ctrl+F5 for integrated browser refresh 🧵 See thread below 👇 🧵 1/3
1
10
56
7,364
Harald Kirschner retweeted
Burke Holland
@burkehollandFeb 18
No shade to other tools. They are all good. There are no bad tools rn. BUT The agent harness for Opus in @code is so good. I keep trying to explain this.
Asish Kumar@asishcodesFeb 17
vs code > cursor i have tested with opus 4.5 opus 4.5 in vs code is way smarter
9
4
1
73
10,422
Next up at @vercel’s Skills.sh launch party, showing how Agent Skills don’t just work in VS Code, but even integrate with extension APIs. Extra bonus for vibing my slides to match the stage lighting.
1
3
1
17
1,147
Next up at @vercel’s Skills.sh launch party, showing how Agent Skills don’t just work in VS Code, but even integrate with extension APIs. Extra bonus for vibing my slides to match the stage lighting.
3
1
10
799
Harald Kirschner retweeted
Asish Kumar@asishcodesFeb 17
vs code > cursor i have tested with opus 4.5 opus 4.5 in vs code is way smarter
7
1
1
20
12,289
Load more