It’s a local, PR-like code review app that works directly with your working tree. You can review committed, staged, unstaged, and untracked changes, leave inline comments, and organize the work into actual reviews without having to push anything anywhere. You connect whatever AI you're using through MCP and have a coherent experience without all the mundane copy-pasting. All comments live in threads attached to specific code changes, so it’s easy to follow what’s going on over time.
Hi all!
I code professionally for the last 15 years - more and more with AI next to me. Being naturally pushed by market to produce more and more output, I now find myself hopping daily between 5 to 7 simultaneous ai sessions.
I realized that the bottleneck is the code review process - and the place where I conduct the reviews.
I work mostly in environments where it's not really an option to dump one shotted AI slop to company's GitHub. By the time my work is pushed there it needs to be more or less ready for code review by my colleagues.
So I found myself juggling between terminals and IDEs copy-pasting my comments about changes created by my agents. I tried many solutions out there, but what I felt all the time, is, that what I need, is my local GitHub
I want that GitHub code review experience but before a commit is made, here on my computer. A single place where all my agents and I, can talk about code changes.
So I build GitWarren.
It’s a local, PR-like code review app that works directly with your working tree. I can review committed, staged, unstaged and untracked changes, leave inline comments, and organize the work into actual reviews without having to push anything anywhere.
What I really like, is that I can just connect whatever AI I use at the moment through MCP and have a coherent experience without this mundane copy-pasting. Also all the comments happen in threads on specific code changes so it's easy to follow what is going on over time.
The app is free, open-sourced, fully local, no accounts, no servers, and you bring your own agents to the party.
On mac you can just:
brew install -cask klarluft/tap/gitwarren
There's also Linux and Windows distribution (Windows is not yet signed - working on it).
I think there must be more people facing similar problems right now, no?
Also, the more I work on this app the more I feel like we may find many other use cases for local GitHub experience - like maybe this can be a meeting place for different agents roaming on your computer, and as an extra, you can also peek through some nice UI what are they doing there - just to stay in the "loop".
Would love to hear how others are handling this.
Love the unique approach to managing your personal workspace. @xfor congrats to your launch!
Report
I run separate git worktrees per agent task rather than multiple agents in the same working tree. Does GitWarren handle reviewing across several worktrees of the same repo at once, or is it scoped to one working tree per instance? If it's the latter, curious if that's a deliberate scope decision or just not the use case you built for first.
@wwenioaguiar once your repository is registered in the app, you define your review compare - 99% of cases it's your feature branch against your main branch - worktree or not it just works out of the box :)
relatable problem, the copy-pasting review comments between terminals is exactly the kind of friction that quietly eats a day. since you mentioned juggling 5-7 simultaneous ai sessions, curious how gitwarren handles it when two of those sessions touch overlapping files in the same working tree at the same time - does it just show the working tree as one flat diff regardless of which agent made which change, or can it flag that two edits are about to collide before you get to the review stage?
@galdayan interesting question - for now, this app is purely about discussing changes on some branches - spotting potential conflicts between these is something I haven't thought about yet to be honest - I need to think about it
@xfor fair answer, honest is better than hand-wavy. it's probably an edge case worth flagging since more people will be running multiple parallel agent sessions on the same repo, not just reviewing sequential changes
Report
Reviewing unstaged and untracked changes without pushing them anywhere is really useful. That feels like a much more natural workflow for catching issues before a PR even exists.
@mathew_james2 thanks for the comment, I often find myself wanting to massage my perfect change to commit. Still improving on it just by using it daily - for example yesterday I found a use case to allow users to open a review from “main” to “main” branch just to see these uncommitted changes. I wanted to push directly to main. In previous release of the app I was blocking reviews pointing to the same branch.
GitWarren
Theneo
Looks very promising, will give it a try this weekend
dif.sh
Love the unique approach to managing your personal workspace. @xfor congrats to your launch!
I run separate git worktrees per agent task rather than multiple agents in the same working tree. Does GitWarren handle reviewing across several worktrees of the same repo at once, or is it scoped to one working tree per instance? If it's the latter, curious if that's a deliberate scope decision or just not the use case you built for first.
GitWarren
@wwenioaguiar once your repository is registered in the app, you define your review compare - 99% of cases it's your feature branch against your main branch - worktree or not it just works out of the box :)
Dial
relatable problem, the copy-pasting review comments between terminals is exactly the kind of friction that quietly eats a day. since you mentioned juggling 5-7 simultaneous ai sessions, curious how gitwarren handles it when two of those sessions touch overlapping files in the same working tree at the same time - does it just show the working tree as one flat diff regardless of which agent made which change, or can it flag that two edits are about to collide before you get to the review stage?
GitWarren
@galdayan interesting question - for now, this app is purely about discussing changes on some branches - spotting potential conflicts between these is something I haven't thought about yet to be honest - I need to think about it
Dial
@xfor fair answer, honest is better than hand-wavy. it's probably an edge case worth flagging since more people will be running multiple parallel agent sessions on the same repo, not just reviewing sequential changes
Reviewing unstaged and untracked changes without pushing them anywhere is really useful. That feels like a much more natural workflow for catching issues before a PR even exists.
GitWarren