This looks well-built, but have you considered using Jujutsu (jj) instead? It elegantly solves the snapshot/checkpoint problem without needing MCP servers or a separate app.
With jj, every file change is automatically captured (no manual commits needed), and you can create lightweight "sandbox" revisions for each Claude Code task. When things go wrong, `jj undo` instantly reverts to any previous state. The operation log tracks everything, making it virtually impossible to lose work.
The workflow becomes: let Claude Code generate messy experimental code → use `jj squash`/`jj split` to shape clean commits afterward. You get automatic checkpointing plus powerful history manipulation in one tool.
I've been using jj with Claude Code for months and it's transformed how I work with coding agents - no fear of breaking things because everything is instantly reversible. The MCP integration seems like added complexity when jj's native capabilities already handle the core problem.
This is something I really wish was just built-in to Claude Code. I want it built in because I don't want to have to think about it beforehand. I should be able to jump back in conversation history and have the state of the code jump back with me, so it's restored to the same state it originally was at that point in the conversation.
(There does also need to be a way to jump back in the conversation history without reverting the code, there are times that is useful too!)
A few weeks ago I asked gemini cli to do something pretty simple and it ran for like 12 minutes and then failed with an exception. Haven't tried it again since.
Cline gives you the ability to jump back to any point in the task. The three options are "Restore task", "restore files" and "restore task and files"
A common experience with these tools is that if you realize you want to change the direction you're heading, it's better to jump back to that point in the work and redo it than it is to try to redirect the tool from where you are. Here's a great post about it on the Cline blog
Hi, the developer here. Already thinking about a way to add it as a background task that can communicate with multiple instances at once. As long as its part of CLAUDE.md, every new project would have it automatically included. Not part of Claude Code, but a good way closer?
You can press ESC-ESC in Claude code to open the conversation history view and go back to a previous point in the conversation. It doesn't restore code, which is still a miss, but it is built in and useful!
This was a pain point in coming from Aider to CC. How to have diffs of the changes once CC has done the changes? Having git commits done the way Aider does it would have saved me a lot of time.
The claude models are just a part of Claude code. I've worked with both copilot with the Claude models and Claude code itself. Claude code is way more capable, and has a greater likelihood of successfully completing a task.
I tend to have auto-accept on for edits, and once Claude is done with a task I'll just use git to review and stage the changes, sometimes commit them when it's a logical spot for it.
I wouldn't want to have Claude auto-commit everything it does (because I sometimes revert its changes), nor would I want to YOLO it without any git repo... This seems like a nice tool, but for someone who has a very different workflow.
"Checkpoints for Claude Code" use git under the hood, but stored in .claudecheckpoints folder, to not mess with your own git. Add itself to .gitignore.
It auto commits with a git message for the changes done through MCP locally.
As someone who doesn't use CC, auto-commit seems like it would be the easiest way to manage changes. It's easy enough to revert or edit a commit if I don't like what happened.
It's also very easy to throw away actual commits, as long as you don't push them (and even then not so difficult if you're in a context where force-pushing is tolerable).
True, but it's harder to reject changes in one file, make a quick fix, etc. I like to keep control over my git repo as it's a very useful tool for supervising the AI.
Yeah I basically have Claude commit via git regularly and the majority of the other features described her can be done via git. I agree it's a neat idea for someone though.
What business? This seems to be completely free, with no pricing, in-app purchases, or anything. That being said, it's strange that it doesn't seem to be open-source.
I don't know what this is but isn't git enough? Incidentally I'm not convinced in my day-to-day for "jujitsu" (jj) but from what I understand about how it works, I've been wanting to give it a try for agent-based coding, based on the way it defaults to saving everything and letting you sort it out after. I do like how Aider commits everything so you can easily roll back, although it ends up with a few too many commits imho.
I've been wanting to experiment also with getting an agent to go back and rebase history, rewrite commits etc in the context of where the project ended up, to make a more legible history, but I don't know if that's doable, or even all that useful.
Git won't catch new files the agent is adding. To get around that you can of course always add all new files, but then you'll potentially have your repo polluted with a bunch of temporary scratch files instead.
You can typically go back and edit git history. But it will require force push and breaking changes. And a few sacrifices to ensure that it doesn't make a mistake because then your repo is potentially broken.
Best way to do that is probably to have it work on branches and then squash merge those.
Another problem I inadvertently dodged by using Jujutsu with Claude Code :)
I tend to send a lone "commit" message to Claude when I think I'm in a spot I may want to return to in the future, in case the current path doesn't work out. Then Claude commits it with a decent message. It knows how to use jj well enough for most things. Then it's really easy to jj new back to a previous change and try again.
Yup, that's what I do. Even for personal projects, with the flurry of changes Claude/other AI assistants make, a branch makes it easier for me to compare changes.
Often I have a branch with multiple commits on it, with each commit corresponding to a message in a conversation with AI on Cursor trying to get a new feature built.
In the end, I can diff the branch against the main branch, and see the sum total of changes the AI agent has made.
Maybe edit/improve manually on my own afterwards. And then, merge.
I always squash and reorganise the commits from aider. It is however awesome that everything is in git directly from the agent. I can’t imagine why not all these tools do this!
I just commit with a “wip!”-prefaced message whenever the LLM pauses and says it’s finished, including new files. You can squash and cleanup later, or revert back to a state before it screwed up.
Also doubles as a way to cohesively look at the changes it made without all the natural language and recursive error/type fixing it does while working.
I don’t understand why people are making it so complicated. You’re saving a minute per iteration with the LLM, tops, at risk of losing control or introducing hard to find issues. It is the definition of diminishing returns.
I don't think jujutsu woild help with this use case -- jujutsu will not save everything because it is not running constantly on your repo. It snapshots the working tree only when you run a `jj` command. Ineffective if an agent is doing work.
I recently started using Aider and had that thought about too many commits. What I realized though was: (1) if I'm going to contribute to a project, I should be working in a local branch and interactively rebasing to clean up my history anyway (and of course carefully reviewing Aider's work first) and (2) if I'm working on my own thing WITHOUT LLM, I tend to prefer to commit every dang little change anyway, I just don't remember to do it because I'm in the zone and then inevitably wish I had at some point.
> I tend to prefer to commit every dang little change anyway, I just don't remember to do it because I'm in the zone and then inevitably wish I had at some point.
That’s what I do too until I developed a practice to break up into thematic commits as I realize I need them. And if I don’t, then I just git reset to the beginning and use git gui to commit lines and chunks that are relevant for a given piece of work. But with experience, I barely do the break down completely - I generally don’t even bother creating commits until I have a starting sense of what the desired commit history should be.
Coding with LLM’s has taught me that there is value to treating AI produced code as a slot machine, that sometimes produces brilliant work and sometimes causes huge bugs. The best way to use LLM’s is to structure the project to maximize the benefit from the rare brilliances and minimize the risk and hazard of those bugs. Version control
has to be treated somewhat like a bonsai, a cycle of growth and pruning with an ultimate framework in mind.
Yeah this is my problem with building my cathedral up front in a prompt, I just don’t work that way. I need to see some interactions and mess with it to tell how it should work.
T shirt estimation doesn't make any sense for AI dev, not one bit. They get epic long features done in hours, and all the shirt sizing comes from cases where the agent circles the drain and needs to be guided, which isn't predictable.
The shirt sizes now are for manual acceptance testing.
My experience with AI tooling is that while it's really useful and great - I don't think i've ever seen a LLM complete an epic long feature well full stop.
Don't get me wrong, it's definitely improved my workflow and efficiency, but you must be winning at roulette if the model is performing well on anything that can't be googled and implemented witihn a similar amount of time.
unless it's claude, where even simple styling changes seem to become epics just when it wants to spit out an extra few thousands lines of code
If you go back and forth with chatgpt/gemini on architectural details first, then get chatgpt to produce a hyper detailed spec (like, almost a program claude can execute), you can get claude to run for 2-3 hours at a shot (particularly with a hook to prevent early stopping). Require >85% test coverage, and bake very clear e2e test paths into the spec, and Claude can come surprisingly close to one shotting big things.
Fair enough! I'll need to give it a try - I tend to mostly use these agents as idea testers as have found them limiting beyond those concepts, but it sounds like they may be quite useful using this spec, thanks!
> T shirt estimation doesn't make any sense for AI dev
It doesn't make sense for NI (natural intelligence) dev, either. Even SCRUM doesn't make much sense. The only Agile thing that really makes sense is Kanban, which is actually known to computer science as dispatch queue.
In the 60s, OS researchers spent time figuring out how to optimally schedule resources for computation. Today, almost nobody uses these techniques. (This is known as "waterfall" in PM parlance.)
It turns out, the cheapest way to schedule computing resources is a simple dispatch queue. Why spend extra time figuring out in what order things need to be done, or how long they will take, if they need to be done anyway? It never made sense and it doesn't matter whether the agent is NI or AI.
Not to trivialize the work being done here but isn’t this as simple as a hook on edit and write tool calls that commits to git? I’m not sure I see the need for a whole app around this vs just the standard git workflow
Hi, the developer here. Its a very early version so there could be a lot of bugs, but I like to use it myself (already found several bugs and updated version soon on its way).
Switching from Cursor to Claude Code this was the biggest loss. Have tried to improve on the Cursor functionality, with features I missed.
I would love any feedback on what you are missing etc
Hey, so it's a bit obvious you vibe coded this, which makes me not want to trust it. Some red flags:
- The Apple icon is a literal apple and not the Apple logo.
- You've got 2 Mac download buttons that do the same thing right at the top, surely one of those is a mistake.
- "Watch it in action" is positioned poorly and fails to be a header for the video. Too close to the button above it.
- "Automatic version control" is not what a checkpoint is? "Version control" means git to almost everyone.
- Privacy link is a fake placeholder.
- "See It In Action" looks like you meant to add images and just forgot?
- You named this like 5 things. The website title is "Checkpoints for Claude Code", the domain is "Claude Checkpoints", the UI website title is just "Checkpoints" as if its a standalone brand, the contact email link uses "checkpoints-app.com", and finally you call it "Claude Diff" in the App Store description. Oh and the HN submission is a 6th one, "Claude Code Checkpoints".
I don't think you succeeded. The product name in the article title is 'Claude Code Checkpoints', and the url is 'claude-checkpoints.com'. Nowhere do you note that Claude is a trademark of Anthropic, or disclaim association with Anthropic, or describe that you obtained permission to use their trademarked name in your product. I personally was confused about whether this was an official product at first. I'd be surprised if Anthropic didn't get peeved about this.
Under the hood, is this simply checkpointing the files in the claude target folder or are you also checkpointing the claude context? One of my biggest pain points is after a few compactions/edits to claude.md and all of a sudden Claude has made a few mistakes and all the context window cruft of fixes it attempted and reverted actually seem to confuse it further and it would be nice to reset to a known happy place code & contextually and retry from there.
to my `~/.claude/CLAUDE.md` file. This has in my experience been enough to get atomic commits after every minor change. I'm probably not the target audience for this though.
Interesting to watch the explosion of projects, even whole startups, which are just a feature addition to Claude Code. Shows how beloved it has become.
Great idea, but i've set it up and the app is pretty unusable for me, there is some sort of blocking process which runs every few seconds and freezes the UI, so you can interact with it properly
Some of the user interaction borders on "disaster" IMO. One puts up with it because it's not a show-stopper for the core value proposition of the software (an LLM agent completing tasks for you), and the core value proposition of the software is really valuable.
The noticeable issues are (1) unpredictable scrolling of the terminal window and (2) a super-buggy text box for inputting the prompt.
In particular if I mash the arrow key too fast while moving around and editing the prompt CC and my terminal's idea of where the cursor is get out of sync somehow and it's tricky to get them re-aligned, and I can't actually input text until I do. The vim mode lets me bypass this but it has its own bugs and is missing a ton of features that I expect. Visual selection in particular seems to be missing? Not entirely certain what things I'm used to are stock vim features vs Spacemacs features but I'm pretty sure visual mode is the former. Regardless, only the very basics seem to actually work. "w", "b", "e", "cw/b/e", "dw/b/e", "esc/i".
So for the most part I actually just edit CC prompts in emacs and paste them.
I resort to this workaround because I am very motivated to use Claude Code. For a less-useful piece of software I would probably just give up.
I've built terminal applications and when not using a dedicated alternate buffer, things like multiline text input and navigation are so easy to screw up. Not to mention when you have to do all the tricks to properly detect key strokes, pastes, etc. It's a mess of printing special codes and carriage returns.
I'm guessing they're using abstraction of some sort, but imo they've done a lot of great features and definitely usable.
That being said- they could just build / use something more like a jupyter notebook and have a wildly more stable and rich experience. Or a classic tui app, but pros and cons.
> That being said- they could just build / use something more like a jupyter notebook and have a wildly more stable and rich experience.
Right, part of the reason it stands out is that we're conditioned to much more functional text input in claude.ai (or competing web apps like ChatGPT).
I assume part of the motivation for the terminal app concept is that all the tool calls run in a deterministic environment (whatever was the environment of the shell where you launched "claude"). A Jupyter-type approach would really muddle up that whole picture (at least from a user perspective).
I really love Claude Code but it's wild to me if
others aren't seeing this.
Is Ctrl+R usable at all? I've given up on it, the whole screen just starts scrolling madly most of the time. Not that I have to press Ctrl+R to get that bug to happen, it's just the most reliable way to do so.
And I've had the input box stuck not accepting input or not allowing me to delete past a certain point a hundred times. By now I know how to get it unstuck (although I couldn't tell you - my fingers figured it out but my brain doesn't know).
While disaster is strong language, Claude Code isn't really a well engineered product, they're just kinda trying shit, they don't have a clear long term vision. The core prompts and agent loop are good though, it's too bad it's not open source so someone could implement them in a client with good UX/engineering (at least without disassembling claude code and being legally questionable).
You can extract prompts with mitmproxy/netcat, and AFAIK there isn't much more to it (bash and todo list are all you need in terms of tools), there's already a lot of simpler tools with better ux:
- sst/opencode and charmbracelet/crush -- related "cc clones" with top tier UX; opencode has near feature parity with cc, crush is more barebones
- block/goose -- a lot of multi-model features and extensions (it's practically a framework), but UI is pretty basic
- antinomyhq/forge -- similar to goose, but last week they merged some PRs with agent-agent communication, yet to see how it works out
- openai/codex, gemini-cli -- both somehow don't even have a way to resume a conversation
- avante.nvim with mcphub.nvim -- neovim plugin that emulates cursor to a degree; has a crazy good hack that makes even older models like gpt4.1 "more agentic" -- it keeps reprompting the model with "STFU and write code" until the model calls a "task_completed" tool; gets diagnostics, formatting and anything else neovim can do "for free"
For the sake of completeness, closed-source:
- amp-cli -- absolutely barebones, zero configuration (they even decide what model you're using for you); one problem -- closed source, no BYOK or subscription, pay per token only
- cursor-cli -- atm unusable, can't even set a global context file
- codebuff -- yet to try it myself, but they have some sort of an overengineered setup with 5+ different models (resoner/coder/file picker (!)/fast apply/...), curious to see how it works in practice (I'm assuming this setup is strictly worse than a single sonnet4/gpt5, but much cheaper)
Claude does have a lot of unique/rare (for now) features -- hooks, sub-agents, background jobs, planning mode, per-prompt reasoning effort controls, executable bash in slash commands.
Only half of them are really useful IMHO, but I wouldn't know that if they didn't have them.
I replied to a sibling comment with my observations - the upshot is the actual user interaction is quite buggy in my experience.
If you typically compose prompts in a separate editor and paste them in you aren't likely to even notice. But it's the kind of thing that would drive me up the wall in a piece of software whose primary function was less impressive.
People don't seriously still "lose work" do they? The last time that happened to me was a floppy disk failure in the 90s... Using LLMs doesn't preclude using git and keeping backups like everyone should.
Am I the only one who is getting tired of all these LLM generated landing pages with their hallmark indigo backgrounds/gradients, unnecessary and tasteless transitions, and meaningless marketing sell points?
"Please don't complain about tangential annoyances—e.g. article or website formats, name collisions, or back-button breakage. They're too common to be interesting."
I thought the same thing when I saw this! Great tool but clearly generated by LLM (Claude code?). It’s interesting that it’s so obvious. I wonder if human taste will evolve to dislike styles that are LLM generated.
I'm not sure what LLM has to do with it. A zillion boilerplate landing pages exist out there regardless and a lot of them look/behave just like that. Copying is copying.
This looks well-built, but have you considered using Jujutsu (jj) instead? It elegantly solves the snapshot/checkpoint problem without needing MCP servers or a separate app.
With jj, every file change is automatically captured (no manual commits needed), and you can create lightweight "sandbox" revisions for each Claude Code task. When things go wrong, `jj undo` instantly reverts to any previous state. The operation log tracks everything, making it virtually impossible to lose work.
The workflow becomes: let Claude Code generate messy experimental code → use `jj squash`/`jj split` to shape clean commits afterward. You get automatic checkpointing plus powerful history manipulation in one tool.
I've been using jj with Claude Code for months and it's transformed how I work with coding agents - no fear of breaking things because everything is instantly reversible. The MCP integration seems like added complexity when jj's native capabilities already handle the core problem.
For anyone interested in the jj + agent workflow, read my post: https://slavakurilyak.com/posts/use-jujutsu-not-git
That said, respect for shipping something that solves a real pain point! The UI looks clean and the diff viewer is nice.
This is something I really wish was just built-in to Claude Code. I want it built in because I don't want to have to think about it beforehand. I should be able to jump back in conversation history and have the state of the code jump back with me, so it's restored to the same state it originally was at that point in the conversation.
(There does also need to be a way to jump back in the conversation history without reverting the code, there are times that is useful too!)
Gemini cli already got this, it's disabled by default.
Ref: https://github.com/google-gemini/gemini-cli/blob/main/docs/c...
if only gemini cli could edit files without getting stuck in an infinite loop
A few weeks ago I asked gemini cli to do something pretty simple and it ran for like 12 minutes and then failed with an exception. Haven't tried it again since.
I asked it go modify a maven pom.xml to support grpc. forgot about it and blew through my usage.
Cline gives you the ability to jump back to any point in the task. The three options are "Restore task", "restore files" and "restore task and files"
A common experience with these tools is that if you realize you want to change the direction you're heading, it's better to jump back to that point in the work and redo it than it is to try to redirect the tool from where you are. Here's a great post about it on the Cline blog
https://cline.bot/blog/how-i-learned-to-stop-course-correcti...
Hi, the developer here. Already thinking about a way to add it as a background task that can communicate with multiple instances at once. As long as its part of CLAUDE.md, every new project would have it automatically included. Not part of Claude Code, but a good way closer?
You can press ESC-ESC in Claude code to open the conversation history view and go back to a previous point in the conversation. It doesn't restore code, which is still a miss, but it is built in and useful!
Cursor has had this for a long time. One of the reasons I don’t switch away.
This was a pain point in coming from Aider to CC. How to have diffs of the changes once CC has done the changes? Having git commits done the way Aider does it would have saved me a lot of time.
You can prompt claude code to do that, or use hooks. You can run lint/test/commit automatically on agent stop.
Yes, but the convenience of aider’s /undo might take quite some work to replicate.
Github Copilot in vscode can do that and also has the Claude models available
The claude models are just a part of Claude code. I've worked with both copilot with the Claude models and Claude code itself. Claude code is way more capable, and has a greater likelihood of successfully completing a task.
I tend to have auto-accept on for edits, and once Claude is done with a task I'll just use git to review and stage the changes, sometimes commit them when it's a logical spot for it.
I wouldn't want to have Claude auto-commit everything it does (because I sometimes revert its changes), nor would I want to YOLO it without any git repo... This seems like a nice tool, but for someone who has a very different workflow.
"Checkpoints for Claude Code" use git under the hood, but stored in .claudecheckpoints folder, to not mess with your own git. Add itself to .gitignore. It auto commits with a git message for the changes done through MCP locally.
As someone who doesn't use CC, auto-commit seems like it would be the easiest way to manage changes. It's easy enough to revert or edit a commit if I don't like what happened.
It's also very easy to throw away unstaged changes, and to stage exactly what you want. I treat the staging process ("git add") as a code review.
It's also very easy to throw away actual commits, as long as you don't push them (and even then not so difficult if you're in a context where force-pushing is tolerable).
True, but it's harder to reject changes in one file, make a quick fix, etc. I like to keep control over my git repo as it's a very useful tool for supervising the AI.
Yeah I basically have Claude commit via git regularly and the majority of the other features described her can be done via git. I agree it's a neat idea for someone though.
This is fantastic, but a very spicy business. All Claude needs to do is add one feature and you've got your whole business nuked from orbit.
What business? This seems to be completely free, with no pricing, in-app purchases, or anything. That being said, it's strange that it doesn't seem to be open-source.
Kind of like Cursor? Just a couple more features (now)
The main benefit of Cursor is avoiding model lock-in. So a solution made only for Claude code seems to be on a much shakier footing.
MCP is widely adopted. This will work with anything that talks it.
I don't know what this is but isn't git enough? Incidentally I'm not convinced in my day-to-day for "jujitsu" (jj) but from what I understand about how it works, I've been wanting to give it a try for agent-based coding, based on the way it defaults to saving everything and letting you sort it out after. I do like how Aider commits everything so you can easily roll back, although it ends up with a few too many commits imho.
I've been wanting to experiment also with getting an agent to go back and rebase history, rewrite commits etc in the context of where the project ended up, to make a more legible history, but I don't know if that's doable, or even all that useful.
Git won't catch new files the agent is adding. To get around that you can of course always add all new files, but then you'll potentially have your repo polluted with a bunch of temporary scratch files instead.
You can typically go back and edit git history. But it will require force push and breaking changes. And a few sacrifices to ensure that it doesn't make a mistake because then your repo is potentially broken.
Best way to do that is probably to have it work on branches and then squash merge those.
> Git won't catch new files the agent is adding.
Another problem I inadvertently dodged by using Jujutsu with Claude Code :)
I tend to send a lone "commit" message to Claude when I think I'm in a spot I may want to return to in the future, in case the current path doesn't work out. Then Claude commits it with a decent message. It knows how to use jj well enough for most things. Then it's really easy to jj new back to a previous change and try again.
You could also create a branch, save your intermediates there, and then squash merge when you are ready.
Yup, that's what I do. Even for personal projects, with the flurry of changes Claude/other AI assistants make, a branch makes it easier for me to compare changes.
Often I have a branch with multiple commits on it, with each commit corresponding to a message in a conversation with AI on Cursor trying to get a new feature built.
In the end, I can diff the branch against the main branch, and see the sum total of changes the AI agent has made.
Maybe edit/improve manually on my own afterwards. And then, merge.
Exactly. This is precisely what I do.
I always squash and reorganise the commits from aider. It is however awesome that everything is in git directly from the agent. I can’t imagine why not all these tools do this!
uhh... .gitignore solves that right?
Yea but then you need need to commit each iteration of each file right away, even if its not working.
Its fine if you just rebase at the end manually, but not good if you don't, your history will be cluttered and as hard to read as the codebase.
Eventually most people who use coding tools will have low knowledge of what is being generated and then they probably never rebase either...
Yup, just rebase.
I just commit with a “wip!”-prefaced message whenever the LLM pauses and says it’s finished, including new files. You can squash and cleanup later, or revert back to a state before it screwed up.
Also doubles as a way to cohesively look at the changes it made without all the natural language and recursive error/type fixing it does while working.
I don’t understand why people are making it so complicated. You’re saving a minute per iteration with the LLM, tops, at risk of losing control or introducing hard to find issues. It is the definition of diminishing returns.
I don't think jujutsu woild help with this use case -- jujutsu will not save everything because it is not running constantly on your repo. It snapshots the working tree only when you run a `jj` command. Ineffective if an agent is doing work.
You can use Claude hooks ( https://docs.anthropic.com/en/docs/claude-code/hooks#stop-an... ) to trigger jj / git automatically.
I recently started using Aider and had that thought about too many commits. What I realized though was: (1) if I'm going to contribute to a project, I should be working in a local branch and interactively rebasing to clean up my history anyway (and of course carefully reviewing Aider's work first) and (2) if I'm working on my own thing WITHOUT LLM, I tend to prefer to commit every dang little change anyway, I just don't remember to do it because I'm in the zone and then inevitably wish I had at some point.
> I tend to prefer to commit every dang little change anyway, I just don't remember to do it because I'm in the zone and then inevitably wish I had at some point.
That’s what I do too until I developed a practice to break up into thematic commits as I realize I need them. And if I don’t, then I just git reset to the beginning and use git gui to commit lines and chunks that are relevant for a given piece of work. But with experience, I barely do the break down completely - I generally don’t even bother creating commits until I have a starting sense of what the desired commit history should be.
Coding with LLM’s has taught me that there is value to treating AI produced code as a slot machine, that sometimes produces brilliant work and sometimes causes huge bugs. The best way to use LLM’s is to structure the project to maximize the benefit from the rare brilliances and minimize the risk and hazard of those bugs. Version control has to be treated somewhat like a bonsai, a cycle of growth and pruning with an ultimate framework in mind.
i love the process of AI vibe coding slowly re-inventing the normal structures of software development
how long until we start seeing software products for scrum management and t-shirt size estimation for claude code
introduce waterfall methodology to the LLM!
People are already doing waterfall with their quest for the perfect prompt that will produce a complete, fully functional implementation.
(These people dont realise that there a lot of tradeoffs to be made that pop up during implementation)
Yeah this is my problem with building my cathedral up front in a prompt, I just don’t work that way. I need to see some interactions and mess with it to tell how it should work.
And soon enough it would need a project manager and HR to keep all the agents in line.
I hope the future becomes AI b2b sass agents selling b2b sass to other agents, just tools buying tools that in turn buy other tools ad infinitum
T shirt estimation doesn't make any sense for AI dev, not one bit. They get epic long features done in hours, and all the shirt sizing comes from cases where the agent circles the drain and needs to be guided, which isn't predictable.
The shirt sizes now are for manual acceptance testing.
My experience with AI tooling is that while it's really useful and great - I don't think i've ever seen a LLM complete an epic long feature well full stop.
Don't get me wrong, it's definitely improved my workflow and efficiency, but you must be winning at roulette if the model is performing well on anything that can't be googled and implemented witihn a similar amount of time.
unless it's claude, where even simple styling changes seem to become epics just when it wants to spit out an extra few thousands lines of code
If you go back and forth with chatgpt/gemini on architectural details first, then get chatgpt to produce a hyper detailed spec (like, almost a program claude can execute), you can get claude to run for 2-3 hours at a shot (particularly with a hook to prevent early stopping). Require >85% test coverage, and bake very clear e2e test paths into the spec, and Claude can come surprisingly close to one shotting big things.
Fair enough! I'll need to give it a try - I tend to mostly use these agents as idea testers as have found them limiting beyond those concepts, but it sounds like they may be quite useful using this spec, thanks!
I found a boost adding LikeC4 to the spec.
> T shirt estimation doesn't make any sense for AI dev
It doesn't make sense for NI (natural intelligence) dev, either. Even SCRUM doesn't make much sense. The only Agile thing that really makes sense is Kanban, which is actually known to computer science as dispatch queue.
In the 60s, OS researchers spent time figuring out how to optimally schedule resources for computation. Today, almost nobody uses these techniques. (This is known as "waterfall" in PM parlance.)
It turns out, the cheapest way to schedule computing resources is a simple dispatch queue. Why spend extra time figuring out in what order things need to be done, or how long they will take, if they need to be done anyway? It never made sense and it doesn't matter whether the agent is NI or AI.
Not to trivialize the work being done here but isn’t this as simple as a hook on edit and write tool calls that commits to git? I’m not sure I see the need for a whole app around this vs just the standard git workflow
Hi, the developer here. Its a very early version so there could be a lot of bugs, but I like to use it myself (already found several bugs and updated version soon on its way). Switching from Cursor to Claude Code this was the biggest loss. Have tried to improve on the Cursor functionality, with features I missed.
I would love any feedback on what you are missing etc
Hey, so it's a bit obvious you vibe coded this, which makes me not want to trust it. Some red flags:
- The Apple icon is a literal apple and not the Apple logo.
- You've got 2 Mac download buttons that do the same thing right at the top, surely one of those is a mistake.
- "Watch it in action" is positioned poorly and fails to be a header for the video. Too close to the button above it.
- "Automatic version control" is not what a checkpoint is? "Version control" means git to almost everyone.
- Privacy link is a fake placeholder.
- "See It In Action" looks like you meant to add images and just forgot?
- You named this like 5 things. The website title is "Checkpoints for Claude Code", the domain is "Claude Checkpoints", the UI website title is just "Checkpoints" as if its a standalone brand, the contact email link uses "checkpoints-app.com", and finally you call it "Claude Diff" in the App Store description. Oh and the HN submission is a 6th one, "Claude Code Checkpoints".
Cool project though, sorry to be so critical.
Are you aware of trademark law?
Yes, tried to make it clear that this is Checkpoints for Claude Code. An addition to Claude Code.
I don't think you succeeded. The product name in the article title is 'Claude Code Checkpoints', and the url is 'claude-checkpoints.com'. Nowhere do you note that Claude is a trademark of Anthropic, or disclaim association with Anthropic, or describe that you obtained permission to use their trademarked name in your product. I personally was confused about whether this was an official product at first. I'd be surprised if Anthropic didn't get peeved about this.
Under the hood, is this simply checkpointing the files in the claude target folder or are you also checkpointing the claude context? One of my biggest pain points is after a few compactions/edits to claude.md and all of a sudden Claude has made a few mistakes and all the context window cruft of fixes it attempted and reverted actually seem to confuse it further and it would be nice to reset to a known happy place code & contextually and retry from there.
Now it’s the files. Have been prototyping about keeping the context also, for an upcoming version.
I just add
> always git commit with a one line message
to my `~/.claude/CLAUDE.md` file. This has in my experience been enough to get atomic commits after every minor change. I'm probably not the target audience for this though.
Or you could use something like NILFS2 [0]. It would be nice to be able to just turn such a function on and off per folder.
[0] https://docs.kernel.org/filesystems/nilfs2.html
Interesting to watch the explosion of projects, even whole startups, which are just a feature addition to Claude Code. Shows how beloved it has become.
Great idea, but i've set it up and the app is pretty unusable for me, there is some sort of blocking process which runs every few seconds and freezes the UI, so you can interact with it properly
Strange, tested on MacBook Pro M3, Pro Intel and MacBook Air M1. All with different OS versions. Straight after the MCP connection setup this happens?
If Claude Code LLMs are reportedly so good, then why is the Claude Code CLI such a buggy, messy, featureless disaster? Are they dogfooding?
Eesh, a disaster? I use it daily to do some pretty impressive things. Wouldn’t call it a disaster.
Some of the user interaction borders on "disaster" IMO. One puts up with it because it's not a show-stopper for the core value proposition of the software (an LLM agent completing tasks for you), and the core value proposition of the software is really valuable.
The noticeable issues are (1) unpredictable scrolling of the terminal window and (2) a super-buggy text box for inputting the prompt.
In particular if I mash the arrow key too fast while moving around and editing the prompt CC and my terminal's idea of where the cursor is get out of sync somehow and it's tricky to get them re-aligned, and I can't actually input text until I do. The vim mode lets me bypass this but it has its own bugs and is missing a ton of features that I expect. Visual selection in particular seems to be missing? Not entirely certain what things I'm used to are stock vim features vs Spacemacs features but I'm pretty sure visual mode is the former. Regardless, only the very basics seem to actually work. "w", "b", "e", "cw/b/e", "dw/b/e", "esc/i".
So for the most part I actually just edit CC prompts in emacs and paste them.
I resort to this workaround because I am very motivated to use Claude Code. For a less-useful piece of software I would probably just give up.
I've built terminal applications and when not using a dedicated alternate buffer, things like multiline text input and navigation are so easy to screw up. Not to mention when you have to do all the tricks to properly detect key strokes, pastes, etc. It's a mess of printing special codes and carriage returns.
I'm guessing they're using abstraction of some sort, but imo they've done a lot of great features and definitely usable.
That being said- they could just build / use something more like a jupyter notebook and have a wildly more stable and rich experience. Or a classic tui app, but pros and cons.
> That being said- they could just build / use something more like a jupyter notebook and have a wildly more stable and rich experience.
Right, part of the reason it stands out is that we're conditioned to much more functional text input in claude.ai (or competing web apps like ChatGPT).
I assume part of the motivation for the terminal app concept is that all the tool calls run in a deterministic environment (whatever was the environment of the shell where you launched "claude"). A Jupyter-type approach would really muddle up that whole picture (at least from a user perspective).
I really love Claude Code but it's wild to me if others aren't seeing this.
Is Ctrl+R usable at all? I've given up on it, the whole screen just starts scrolling madly most of the time. Not that I have to press Ctrl+R to get that bug to happen, it's just the most reliable way to do so.
And I've had the input box stuck not accepting input or not allowing me to delete past a certain point a hundred times. By now I know how to get it unstuck (although I couldn't tell you - my fingers figured it out but my brain doesn't know).
This is a funny comment because it's just made up
While disaster is strong language, Claude Code isn't really a well engineered product, they're just kinda trying shit, they don't have a clear long term vision. The core prompts and agent loop are good though, it's too bad it's not open source so someone could implement them in a client with good UX/engineering (at least without disassembling claude code and being legally questionable).
> not opensource
You can extract prompts with mitmproxy/netcat, and AFAIK there isn't much more to it (bash and todo list are all you need in terms of tools), there's already a lot of simpler tools with better ux:
- sst/opencode and charmbracelet/crush -- related "cc clones" with top tier UX; opencode has near feature parity with cc, crush is more barebones
- block/goose -- a lot of multi-model features and extensions (it's practically a framework), but UI is pretty basic
- antinomyhq/forge -- similar to goose, but last week they merged some PRs with agent-agent communication, yet to see how it works out
- openai/codex, gemini-cli -- both somehow don't even have a way to resume a conversation
- avante.nvim with mcphub.nvim -- neovim plugin that emulates cursor to a degree; has a crazy good hack that makes even older models like gpt4.1 "more agentic" -- it keeps reprompting the model with "STFU and write code" until the model calls a "task_completed" tool; gets diagnostics, formatting and anything else neovim can do "for free"
For the sake of completeness, closed-source:
- amp-cli -- absolutely barebones, zero configuration (they even decide what model you're using for you); one problem -- closed source, no BYOK or subscription, pay per token only
- cursor-cli -- atm unusable, can't even set a global context file
- codebuff -- yet to try it myself, but they have some sort of an overengineered setup with 5+ different models (resoner/coder/file picker (!)/fast apply/...), curious to see how it works in practice (I'm assuming this setup is strictly worse than a single sonnet4/gpt5, but much cheaper)
Claude does have a lot of unique/rare (for now) features -- hooks, sub-agents, background jobs, planning mode, per-prompt reasoning effort controls, executable bash in slash commands.
Only half of them are really useful IMHO, but I wouldn't know that if they didn't have them.
By which metric is it a disaster? Seems quite solid and impressive to me as it is.
I replied to a sibling comment with my observations - the upshot is the actual user interaction is quite buggy in my experience.
If you typically compose prompts in a separate editor and paste them in you aren't likely to even notice. But it's the kind of thing that would drive me up the wall in a piece of software whose primary function was less impressive.
Git add doesn't work?
I have a strong suspicion that if you need, or feel you need something like this, you're misusing coding agents.
I want to do this with btrfs subvolumes/snapshots.
So basically, "watch jj"
People don't seriously still "lose work" do they? The last time that happened to me was a floppy disk failure in the 90s... Using LLMs doesn't preclude using git and keeping backups like everyone should.
this is coming to Claude Code
They have hinted at it?
Disappointing that Claude’s chat interface is falling so far behind.
- no large context
- no zipfile uploads
- no multi file downloads
Am I the only one who is getting tired of all these LLM generated landing pages with their hallmark indigo backgrounds/gradients, unnecessary and tasteless transitions, and meaningless marketing sell points?
Of course you're not the only one, but:
"Please don't complain about tangential annoyances—e.g. article or website formats, name collisions, or back-button breakage. They're too common to be interesting."
https://news.ycombinator.com/newsguidelines.html
Plus the endless emojis and the cards with images completely missing from the bottom section where you're supposed to "see it in action."
It shows a lack of respect for yourself and your potential readers when you build something like this.
I thought the same thing when I saw this! Great tool but clearly generated by LLM (Claude code?). It’s interesting that it’s so obvious. I wonder if human taste will evolve to dislike styles that are LLM generated.
Yes, I had to do a double take to realize this wasn't from anthropic.
It might be boring but it's easily understandable - I know where to click, where to look etc.
It's kind of a minimalism; it's not pleasing (to me) but it works
I'd argue it's actually the exact opposite of minimalism. Just an abundance of colours, emojis, and marketing jargon. It shows a lack of taste.
I'm not sure what LLM has to do with it. A zillion boilerplate landing pages exist out there regardless and a lot of them look/behave just like that. Copying is copying.
LLMs are known to overuse purple:
https://x.com/adamwathan/status/1953510802159219096
https://ai-engineering-trend.medium.com/the-mystery-behind-a...
Yeah. "LLM" and "Wordpress" is the same.
Yeah definitely feels vibe coded. I am so sick of the overuse of icons.