Burke Holland

@burkeholland • 25,397 subscribers

Working on GitHub Copilot @microsoft @github. My views.

Shorts

22,527 views

82,909 views

37,084 views

22,094 views

14,839 views

Videos

sweetdream.ai

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

Private Show

Join now for exclusive access

Free preview available • Premium content

59,055 views • 10 days ago

241,703 views • 1 month ago

23,448 views • 21 days ago

94,784 views • 4 months ago

124,377 views • 5 months ago

82,970 views • 4 months ago

72,340 views • 4 months ago

12,506 views • 28 days ago

34,563 views • 3 months ago

19,389 views • 1 month ago

10,112 views • 23 days ago

26,819 views • 2 months ago

28,512 views • 3 months ago

79,731 views • 11 months ago

13,147 views • 1 month ago

21,370 views • 3 months ago

48,080 views • 10 months ago

36,152 views • 11 months ago

34,733 views • 1 year ago

13,694 views • 3 months ago

Live Cam

Burke Holland

Shorts

I remember when building a date picker was one of the hardest things you could do in frontend dev. This was just me telling @GitHub Copilot "/grill-me lets make a date picker". It's really an incredible time to be doing this job. Huge thanks to Matt Pocock for grill-me.

Don't put that API key in clear text for your MCP Server! Use the inputs in Visual Studio Code to encrypt the key at rest. This way you can be sure you'll never accidentally make it public.

Engineer who works on the terminal for @Code just showed me that the latest release autocompletes git branch names...🤯 Thanks, Megan!

The integration between the Copilot CLI and Visual Studio Code is actually brilliant. Terminal tab titles are now updated by the agent in realtime and you'll see a 🤖 icon while the agent is working. It's always the little things, ya'll.

Let's do @Code UI/UX tips/tricks all day. Here's another of my faves... Wrap the code blocks in the chat. Makes it much easier to scan the responses.

Videos

Watch Anya Live

I fed this tweet to @GitHub Copilot verbatim and Opus 4.6 said...

The @GitHub Research folks released a "Rubber Duck" agent for the Copilot CLI. Automatically get a review from a model from a different AI family. And their data shows that yes, this does in fact help. Quite a bit.

GitHub just released a technical preview of the "GitHub Copilot App" - a new agentic development tool that brings agents and @GitHub together in one pane of glass. I've been using this daily for a few weeks. I think you're going to love it.

I discovered the wildest feature of GitHub Copilot CLI by complete accident last night. You can have models call each other. So you can have Opus, Codex and Gemini all working together in the same chat.🤯

Opus 4.5 is absolutely goated. I built AN ENTIRE VIDEO EDITING APPLICATION FOR WINDOWS IN ABOUT AN HOUR. No tricks. No special prompts, no prompt driven dev. Just me, Visual Studio Code, and Opus 4.5 having a convo via the mic. I feel like this is what we were promised, and it's here.

When I heard about the Copilot SDK, my first thought was whether or not I could build an agent to control my desktop. So I did that. Being able to build and pass custom tools to the agent is op.

Microsoft released a Work IQ MCP server that lets you basically bring organizational context into anything with simple prompts. So I built a @Code extension to tell me when I'm missing meetings and show me all documents that might be relevant to the current repo. Coming soon!

In the Copilot CLI you can /ask the agent questions while it's busy running other things and it has the context of your session. My new favorite feature.

The Copilot CLI - how to configure and use it, some tweaks I make and a few of my own workflows / custom agent. I was proud of this one. Pretty much a one shot. It's wild how easy it is to make content when a thing is good.

Hooks in Copilot CLI might be the answer to all of our AI problems

The @GitHub Copilot CLI lets you /fork conversations into different sessions you can toggle between. I've been using this lately to go down different roads with the same plan before I commit to one of them.

You can legit build your own OpenClaw🦞 from scratch in about 20 minutes with the @GitHub Copilot SDK. And when you build it, you can make it do anything. Here's how to build a personal AI assistant, and how I'm using one today.

The @GitHub Copilot CLI has a feature called "autopilot" that is essentially its version of a Ralph loop. But how does it hold up to an actual Ralph Loop? Let's find out...

Now Visual Studio Code will generate your copilot-instructions file for you. You will get CRAZY better results by having these instructions for the AI...

2 impressive things happening here... 1. You can now access Copilot CLI sessions running on your machine from your phone 2. There is a remodel happening in my house but you can't hear any of it because of how good @adobe's AI is at filtering it out.

The beautiful thing about the GitHub Copilot CLI is that it is endlessly hackable. You can use it to stand up any workflow you can dream up. Here's one workflow I've been toying with - a whole dev team, managed by Copilot.

4.1 vs Claude "search for the lifx api docs official and create a markdown file that documents the api. include auth, how to call the api, as well as a markdown table of all api endpionts and their parameters. save to the docs folder."

If you're not using Hyperspace, you are absolutely using GitHub the hard way.

MCP broke my brain a little because it rethinks the entire workflow. Debug, diagnose and file the issue with Playwright and GitHub MCP. All from agent mode in Visual Studio Code. What are you using MCP for beyond code?

Both the GitHub Copilot CLI and @Code now support calling custom subagents from the main thread. Which means you can start to get into some pretty interesting orchestration scenarios.