Skip to main content

Why is my AI agent bribing a hamster?

Recently I was watching one of my AI agents chew through a feature, and in the middle of its chain of thought it dropped: "Bribing the hamster." My first thought was that I was hallucinating instead of the AI model. My reaction was to check my coffee. My second was to find out what was actually going on.

Turns out, it's not a hallucination. It's a feature.

The setting behind it

VS Code's Copilot Chat has a setting called chat.agent.thinking.phrases. Instead of a static "Thinking..." or "Working..." label while the agent is doing its multi-step work in the background, it rotates through a list of playful phrases. Here's what that setting looks like in practice:

"chat.agent.thinking.phrases": {
  "mode": "replace",
  "phrases": [
    "Bribing the hamster",
    "Reticulating splines",
    "Untangling the spaghetti"
  ]
}



Remark: "Reticulating splines" is a direct nod to SimCity, where it famously meant nothing at all. Just a fake loading message to keep you entertained while the game did its actual work. GitHub is playing the same joke here.

Why a hamster specifically

The hamster is GitHub's own loading icon. The animal that spins on a wheel while something is running in the background. It's part of a longer lineage of GitHub mascots: the classic "Ship It" squirrel, and now the hard-working loading hamster. Giving it a personality ("bribing" it) is a small wink at the fact that agent mode is still, underneath all the autonomy, something you're waiting on.

Tip: You can override this list yourself. Set chat.agent.thinking.phrases to "mode": "replace" with your own array to swap in your own jokes, or "mode": "append" to add to the existing set instead of replacing it.

The bigger thing hiding behind the joke

The phrase itself is throwaway, but it's a marker of something real: agent mode isn't a single-shot prompt completion anymore. Behind that spinning hamster, Copilot is scanning your workspace, building a multi-file execution plan, running tools, and self-correcting as it goes. The silly loading text exists precisely because that work can take a while, long enough that a static "Thinking..." stopped being good enough UX.

So next time your agent tells you it's bribing a hamster, or reticulating splines, or untangling spaghetti: it's not broken, and neither is your coffee. It's just VS Code keeping you company while the real work happens underneath.

More information

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         ...

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) ...

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...