Skip to main content

Turning a single work session into a reusable Skill with Microsoft's Skill Recorder

In my effort to adopt an 'AI first' philosophy, I spend a lot of time automating my day-to-day activities and transforming them to skills. It's not the most exciting activity: I know exactly how to do a task, I've done it a dozen times by hand, but writing it down as a clean SKILL.md for an agent takes almost as long as just doing the task again.

Microsoft created an open-source a tool that skips that step entirely: Skill Recorder. The idea is simple: you record yourself doing the task once, and it generates the skill for you.

Sounds good? Let’s give it a try…

What it actually does

Skill Recorder is a desktop app (Electron, macOS-first with Windows 11 support) that captures a real work session on your screen: clicks, app and window switches, the pages you visit, clipboard snippets, and optionally your spoken narration. Nothing leaves your machine while you're recording. Capture, storage, frame extraction, and narration transcription all happen locally.

This changes when you hit Analyze. That's when the session gets sent to GitHub's cloud, where the GitHub Copilot CLI reconstructs what you actually did: one overall intent, plus an ordered list of steps. You can review and edit that until it reads right, then generate:

  • a Skill: a SKILL.md procedure an agent runs on demand, or
  • an Automation: the same procedure on a schedule or trigger.

The tool was mainly build to use with Microsoft Scout, Copilot Cowork, or Copilot Studio, and tries to use the agent's native tools over literally replaying your UI clicks. Record yourself submitting one form, and the agent generalizes it into a skill that be used to submit all of them, not just the exact one you recorded.

Installing it

Skill Recorder ships as a source release: one command downloads a pinned Node.js runtime and builds the exact release commit on your machine. Nothing installs globally. The only prerequisite is that you have a GitHub account with Copilot access. The Copilot CLI itself ships with the app.

Grab the commit id from the latest release, then on Windows you can compile and install the app:

$commit="<40-character-release-commit>" 
$env:SKILL_RECORDER_COMMIT=$commit 
irm "https://raw.githubusercontent.com/microsoft/skill-recorder/$commit/install.ps1" | iex

How to use it

  • Record. Hit record (⌘⇧R / Ctrl+Shift+R works from anywhere) and just do the task. A small always-on-top bar shows capture and mic state while you work, and you can mute, switch mics, or discard the take if it goes wrong.


  • Analyze. Click Analyze. This is the point where your event timeline, extracted screen images, and narration text go to GitHub's cloud for Copilot to process.

 


  • Review. You get back an intent statement and an ordered step list. Edit it until it's accurate.

  • Create. From the approved analysis, generate a Skill.


And here is (part of) the exported skill:


Just one warning

Every recording that goes through Analyze is sent to GitHub's cloud, including window and document titles, URLs, clipboard previews, and screen images. Skill Recorder reminds you of this before every recording, and it's not subtle about it: don't record, type, paste, or narrate passwords, tokens, API keys, or other credentials. If your task involves logging into something, do that part outside the recording.

Tip: the narration feature transcribes on-device via Whisper (99 languages, one-time ~250 MB model download), so narrating your intent out loud doesn't add an extra cloud dependency. Only the Analyze step does.

To conclude

The bottleneck in building agent skills has never really been the agent's capability. It's the time it takes us to write down a procedure precisely enough for an agent to follow it, and generally enough that it still works next week when the data looks slightly different. Skill Recorder attacks that bottleneck directly: you do the task once, the way you'd do it anyway, and Copilot handles the abstraction from "what I clicked" to "what I intended."

Too soon to tell if this really is a must have. But worth trying on one of your own repetitive admin tasks before you decide whether it belongs in your agentic toolkit.

More information

Labels:

Popular posts from this blog

Podman– Command execution failed with exit code 125

After updating WSL on one of the developer machines, Podman failed to work. When we took a look through Podman Desktop, we noticed that Podman had stopped running and returned the following error message: Error: Command execution failed with exit code 125 Here are the steps we tried to fix the issue: We started by running podman info to get some extra details on what could be wrong: >podman info OS: windows/amd64 provider: wsl version: 5.3.1 Cannot connect to Podman. Please verify your connection to the Linux system using `podman system connection list`, or try `podman machine init` and `podman machine start` to manage a new Linux VM Error: unable to connect to Podman socket: failed to connect: dial tcp 127.0.0.1:2655: connectex: No connection could be made because the target machine actively refused it. That makes sense as the podman VM was not running. Let’s check the VM: >podman machine list NAME         ...
Read more

Cache stampede: when our cache turned against us

While investigating some performance issues, we ran into an ASP.NET Core API that cached a fairly expensive aggregation query for 60 seconds. Under normal load, that was fine: one request rebuilds the cache, everyone else reads from it. Under peak load, dozens of requests would arrive in that same expiry window, all see a cache miss, and all fire the same expensive query in parallel. The database didn't like that. That was the moment when our caching layer stopped helping and started hurting. A burst of requests comes in at the same time, all miss the cache, and all go hammer the database or the downstream API at once. That's a cache stampede . The cache was supposed to protect our backend, and for a few hundred milliseconds it did the opposite. Why this happens IMemoryCache.GetOrCreate (and its async sibling) looks like it protects you, but it doesn't add any locking on its own. Look at the naive version: public async Task<Report> GetReportAsync(string key) ...
Read more

The role of ActivitySource in OpenTelemetry for .NET

While doing some pair programming to integrate OpenTelemetry tracing to a .NET application, we had a discussion on how to use the ActivitySource . It looks simple. You new one up, give it a name, start an activity, done. The discussion started when we added a second ActivitySource with the exact same name in a different class. This made us wonder: "Are we duplicating traces now? Is this a memory leak? Do we need a singleton?" So we decided to dig deeper. This post is what we learned… What ActivitySource actually is ActivitySource is part of System.Diagnostics , not part of the OpenTelemetry NuGet packages. Microsoft built tracing primitives directly into the BCL, and OpenTelemetry's .NET SDK simply listens to them. This is why you can add distributed tracing to a library without taking a dependency on OpenTelemetry at all. An ActivitySource is a factory for Activity objects, and an Activity is .NET's name for what OpenTelemetry calls a span.(don’t ask m...
Read more