Share of wallet (dollars spent) vs share of tokens used. Anthropic vs OpenAI
Astra = top model by spend last week
Luna = top model by tokens (by a LOT)
Wrote a tiny memo on how to let a sandboxed Claude Code access a local browser using Chrome DevTools MPC via a forwarded debugging port 9222.
It's just a few commands, and the setup is fairly secure 🧙♂️
https://t.co/VIVYBNss4K
Just dropped a new version of my Stitch CLI.
It's full of code management features for your @stitchbygoogle designs.
Give it a try (and a ⭐️) and lmk.
npx @_davideast/stitch-mcp view --projects
We've added a new command to Claude Code called /insights
When you run it, Claude Code will read your message history from the past month. It'll summarize your projects, how you use Claude Code, and give suggestions on how to improve your workflow.
"I don't have a GPU" is officially over.
VS Code now connects directly to Google Colab.
→ You get a free T4 GPU inside your editor.
→ Takes 2 minutes to set up. Their compute.
Cursor tip: before pushing a PR, ask the agent to break your changes into a series of commits for easier review.
You can even stack skills like:
1. /code-review
2. /simplify (can we get the same result with less code?)
3. /deslop (remove unnecessary comments)
4. /commit-pr
As a big fan of Duolingo’s animations, I’ve always loved how they bring motion into their app experience.
So I decided to recreate one of their animations using @rive_app ever since I saw Duolingo using it, I’ve wanted to give it a try. 🎨✨
Everyone uses Docker, but very few people know how debug a Docker container.
No, I’m not talking about "docker logs", or "docker inspect", or "docker exec"
I’m talking about "docker events".
You probably don’t even know it exists.
But it streams low-level, real-time event logs from the Docker daemon; not just your container.
Imagine this:
Your container is restarting in PROD.
But it works on your machine.
So like a real champ, you SSH into the server and start your usual debugging ritual:
👉 docker logs → nothing
👉 docker inspect → static
👉 docker exec → can't connect before it dies
Sounds familiar, right?
What next? Let me guess:
You start tweaking the Dockerfile and doing consecutive docker run attempts.
Because you’ve never heard of docker events.
Now try this instead:
docker events --filter container=<your_container_id>
Output:
2025-05-04T18:20:01Z container create ...
2025-05-04T18:20:02Z container start ...
2025-05-04T18:20:03Z container health_status: unhealthy ...
2025-05-04T18:20:04Z container kill signal=SIGTERM
2025-05-04T18:20:04Z container restart
👉And boom.
Healthcheck failed → container killed → Docker auto-restarted it.
And your app logs?
They don’t show that. Because the healthcheck is outside your main process.
Bonus tips:
- Scope timeframes:
docker events --since 30m --until "2025-05-04T18:00:00"
- Want structured output?
docker events --format '{{json .}}'
- Need a full system timeline during an outage? Just remove the filter:
docker events
And before you tell me:
"Well actually, in modern infra, you should use Grafana, Loki, Prometheus for observability…"
Sure. Good luck debugging:
- OOM kills before metrics are scraped
- Missed volume mounts
- Broken init containers
- Sidecars that silently crash and restart
docker events sees it all. In real time.
If you're debugging Docker without it,
You're not debugging, you're just guessing.
Within two years, you can create a full game like this from scratch in less than a week.
In the meantime, here's the prompt to create the images for a game like this with Midjourney 👇🏼
"Cline for Slides"
We made a .clinerules file for making slides with @Slidevjs -- now Cline can build your decks.
Here's how we turned a YouTube video into a deck: 🧵