In my head, these 2 rules are making sense. The LLM has a smaller context to juggle with, and there is a better separation of concerns.

Meanwhile, I noticed that Claude has a new option after planning. Instead of just bypassing the permission to implement the feature, it also asks to clear the context and calls a new prompt with only the plan it created. This removes the need to create a plan file, like I always did before, and they are trying to reduce the need for the Ralph-Loop. Which I have not tried so far. This is also on my to-do list.

95 of #100DaysToOffload
#log #ai #claude
Thoughts? Discuss...

After a bit of research, I found nothing that resonated with me. Nothing that looked like a good fit. While researching, I found out that there were Sudokus for kids with shapes on paper. We tried them, and he liked it. Then it clicked, and I fired up Jippity. We talked and created a plan for an MVP game in React Native. Thanks to my #AdventOfProgress, I already had experience with React Native and a base from where we can start.

There were points that made the implementation easy. For example: offline first and no auth, so no Supabase or backend dependency is needed and less complexity. A clear and simple scope. No game engine is needed because of the missing game loop and simple graphics.

I had a lot of fun, and it was difficult to stop. After the initial MVP was complete, I kept adding features. AND it is the first project I started without having a domain name first. 🎉

The finishing line is in sight. There is a little bit of work left on the landing page that needs to be done and smaller visual issues inside the app that I need to address. Also preparing all the data for the stores, for example, taking screenshots for each language and adding marketing texts, needs a lot of work. It will be my first release of an app in the stores. It should at least look a bit polished. 😅

87 of #100DaysToOffload
#log #dev #game #kids #ai #claude
Thoughts? Discuss...

Recently, I saw Clawd.bot, which is a personal AI assistant and looks promising. Let's see what use cases I can find here.

84 of #100DaysToOffload
#log #ai #claude #dev
Thoughts? Discuss...

Last weekend, I wanted to try it out for a task where I didn't want to sit next to Claude, watch it, and prevent it from failing. So I’ve made a new attempt. “We” planned the task together; I tried to be as precise as I could be with the prompt and got a first draft as a result. Then we iterated and refined the plan until I was confident with it.

On the weekend, we had planned some family time. It was the best opportunity to let Claude go wild with the planned task. Later in the evening when I was back with the family, I was a little excited about how it worked out. I took a first look, and Claude had completed the task. The result was working, and the tests were green. So this looked promising. I peeked at some files, and it looked good.

After I did a full review on the next day, I was quite happy with the result. There were some small adjustments where Claude was not following the rules exactly, but this is ok. Like checking a PR from another colleague.

Maybe I should try this more. One idea would be that after a finished workday, I go into yolo mode with Claude and review the results on the next day. But before I can do this with confidence, I need to refine my Claude settings. The settings are there, but I never touched them, really. There is potential here. :)

50 of #100DaysToOffload
#log #dev #claude
Thoughts? Discuss...

Writing with CC is like writing with an elaborate sane person, who really knows more than you. (Most of the time.)

When writing a prompt to either of them, I have this mindset above for each in my head, and this is how my prompts look. Maybe this is the reason why my outcome on cursor is not as good as in cc?

Nevertheless is gain something from both of them, so the current state is ok for me.

41 of #100DaysToOffload
#log #dev #claude
Thoughts? Discuss...

Preliminary skirmish

But let’s assume I already wrote the post. I’m really happy with the terminal. I needed to migrate away from Alacritty+Tmux because of the lack of Windows support from Tmux. Overall it was a good decision to migrate. The only thing that currently bothers me is that there is a “bug” where the font rendering misses some characters but brings them back on config reload. (cmd+r)

Here starts the main post

After starting my journey with Claude Code, I needed a handy shortcut to open CC in a 1/3 split pane. Here is the part of my config, which is doing the Job:

config.keys = {
        --...
	{
		key = "c",
		mods = "LEADER|SHIFT",
		action = act.SplitPane({
			direction = "Right",
			command = { args = { "bash", "-lc", "claude" } },
			size = { Percent = 33 },
		}),
	},
	--...
}

You'll hit leader & shift+c and it will open a 1/3 sidebar with Claude Code in it. Done.

For me it is a nice little helper. Hopefully for someone else too. Let me know if this was helpful.

37 of #100DaysToOffload
#log #dev #claude #wezterm
Thoughts? Discuss...

Why are you writing different Python and Bash scripts?

Why are you creating duplicates you then have to remove again?

Claude, I think you are drunk. Lets /clear your brain and start again.

35 of #100DaysToOffload
#log #dev #claude
Thoughts? Discuss...