My status line shows 5h and 7d quota usage among other things like model, context usage, git branch, etc. Unfortunately the whole status line disappears half of the time from CC’s broken ass UI.
Personally I have multiple instances of Claude on my PC dedicated to different areas (split between my private subscription and my work one) and it is useful to see both at a glance. And I don't have to keep the terminal open to see how much time is left for quota to be reset. But for single-Claude users this is indeed a valid solution!
Can we stop with this way of installing things? This is already on brew and it’s a menu bar app. Just let me download it instead of executing some arbitrary bash script
I need to turn this into a blog post at some point:
Some of my early bad experiences with Linux arose because I installed software in ways that broke the system quite impressively. This taught me that with most Linux systems, you are not really supposed to just download random packages and shoe-horn them into your system. Or blindly compile and sudo make install things that could conflict with already-installed software.
The curlpipe pattern feels like a return to YOLO'ing your software installations, like the bad old Windows days where any INSTALL.EXE could overwrite another program's DLLs, wherever they lived on the disk. I trust the developers of my operating system to know what they are doing when they package software for it because most Linux distro communities have a vetting or code review process. I even sometimes trust people and projects who build their own packages for my distro and host them in their own third-party repo. Because that alone shows they probably have learned the bare minimum of things necessary to not break their users' systems.
But a curlpipe script? In my experience, the percentage of developers on GitHub who can write decent Python or Javascript code, and yet don't understand the basic concepts of The Unix Way and how to write safe, portable shell scripts is Very High. I am not going to hand control of my computer over to a random shell script on the Internet, end of story. If your program is any good, provide some generic hand-written instructions on how to build and/or install it, and I will follow those so that I can vet or modify each step as needed. I don't have time in my life to code review your shell script for a project that I was only mildly interested in to begin with.
Agree, and a lot of software on Mac tends to do the same thing: Codex, Claude Code, and so on. Fortunately, it’s just one of many options, and we can use Homebrew in 95% of cases.
P.S. Installing Homebrew is also officially done by running a shell script from GitHub.
Created the same tool some time ago for Codex and Claude code: showing percents for both in the menu bar and detailed stats when you click on it. Using it literally every day, feel free to try: https://github.com/max2697/RateLimited
Feedback greatly appreciated!
Honestly drives me crazy that CC doesnt show this on the bottom status bar a simple percentage “Weekly Quota Left: 99%” would be useful. Also noticed and idk if it was me or some mishap, a lot of my Claude sessions were not autocompacting, maybe its /loops fault not sure, but it made sense why I finally reached my weekly limit so insanely quickly recently.
I just want to know if my next prompt is about to eat shit at the last token. The number of times it like half works on something and falls apart halfway through.
You likely still wouldn't be able to figure that out even if you had a percent usage view available. Will your request take 1K tokens or 10K? Who knows! That's the magic of non-deterministic black boxes.
Doesn't it autocompact only if it is about to run out of context? On these newer models the context is 1M tokens so it's quite difficult to reach. I run compacts manually when I stop in a good place and tokens are over 300k
1m is too high, they used to autocompact a little too soon, maybe I tweaked defaults, it was kind of annoying, 200k tokens is too soon for me, I feel like the sweet spot is around 400k tokens.
Funnily enough Groks API charges less when you are below 400k tokens.
Half a million can get you reasonably far with enough of your codebase within the models context window.
I lowkey blame the 1m context window as the start of Anthropics worse woes earlier this year.
The Claude app has a usage indicator when you switch to code mode. I’ve been using this more and more now that their sandbox functionality has Improved.
https://www.tyleo.com/blog/love-letter-to-the-claude-code-do...