If you've ever pasted ten files into ChatGPT, asked for a fix, then manually copied the result back into ten different files by hand, this post is for you!
Already using ReptClip? Skip the intro and jump straight to what's new in v2.0
The problem
Coding agents like Claude Code and Cursor are great, but they cost money, and not everyone wants (or can afford) a subscription. Meanwhile, free chatbots like ChatGPT and Claude.ai are sitting right there, perfectly capable of writing good code, as long as you can get your project's context in front of them and get the resulting changes back into your files without losing your mind 🤔
That gap, pasting context in and applying diffs back out, is tedious enough that most people just don't bother. They either pay for an agent or they give up and do it manually, file by file.
Enter ReptClip
ReptClip is a free, open source VS Code extension that closes that gap 🙂 It lives right in VS Code's bottom panel, as its own tab next to your terminal, so there's no separate window or webpage to juggle. No API key, no subscription, no vendor lock-in. It works with literally any free chatbot you could find online.
Here's what the whole loop looks like in practice:
Here is the explanation. It's a 3-step process:
- Copy context to clipboard. Pick the files you care about (by path or glob pattern), hit the "Context" button, and ReptClip copies your project structure, AI instructions, and the file contents as clean Markdown straight to your clipboard (each section can be easily toggled on/off).
- Paste into your chatbot. Drop it into ChatGPT, Claude, Gemini, whatever you're using, along with your prompt.
- Apply the diffs. The chatbot replies with its normal explanation, along with a single code block containing Search/Replace diffs. Copy that block, click "Apply Diffs" in the ReptClip panel, and your files are updated!!! 😎
If the chatbot needs to see more of your project mid-conversation, it can ask for it, and you feed that back in with one more click.
Some things that make it safe to actually trust with your codebase:
-
ReptClip only touches files that are tracked by git. Anything your
.gitignorefiles (at any depth) exclude is completely off-limits, for reading, searching, writing, creating, or deleting..gitfolders are always off-limits too. Changing a.gitignorefile always asks for confirmation first, since that can un-ignore something that was previously protected. If something ever goes wrong anyway, you can simply discard the changes with git. - Applying a diff is all or nothing. If one block fails to match, nothing gets written. Fuzzy matching handles whitespace and line-ending differences, but still requires a unique match.
- You get 10 levels of Undo/Redo specifically for applied diffs, separate from your normal editor undo.
- And MUCH more! 🙂 README
The UI state is saved per project automatically, so you don't have to re-enter when you come back or be mixed up with other projects.
How to install
It's available in both VS Code Marketplace and Open VSX. So no problem if you're on a VS Code alternative like Cursor or Windsurf. Simply search "ReptClip" and install.
What's new in v2.0
This is a major version bump. The headline feature is a complete rework of how ReptClip feeds extra context to your chatbot mid-conversation, and it also brings several useful new features for applying diffs 🎉
"More Files" is gone. Say hello to "More Context."
In v1, if the chatbot wanted more files, it would ask for a list of files. That's it. That's all it could ask for.
In v2.0, the chatbot can generate a small command block containing three types of commands, and ReptClip generates the output (no subprocess) and copies it to your clipboard!
-
cat <path>reads a full file. -
grep <flags> <pattern> [paths...]searches your project's contents, ripgrep-style, running in-process (no subprocess spawned). -
read-lines <path> <start> [end]pulls a specific, 1-indexed line range out of a file, handy for huge files where you only need a slice.
So instead of the chatbot blindly asking for whole files it might not even need, it can now search your codebase for the right TODO, grep for every usage of a function, or read just the relevant 20 lines of a 2,000-line file. It supports the flags -i, -w, -F, -g, -m, -v, and more. Everything still respects your .gitignore the same way the "Context" button does.
Click "More Context," and it works exactly the way "More Files" used to: read the clipboard, validate everything up front (so a bad path or invalid pattern fails the whole batch with a clear error instead of a half-built result), then build the output and send it back to your clipboard or output file.
Line counts in the project structure
The "Project structure" section now lists each file with its raw line count next to it, like src/main.py (387). This gives the chatbot a much better sense of file size before it decides whether to cat the whole thing or just read-lines a chunk of it. You can turn this off with the new reptclip.showLineCounts setting if you'd rather keep the structure listing minimal.
Move, rename, and explicit delete in diffs
The diff format your chatbot generates picked up two new one-line operations:
-
old/path.txt -> new/path.txtmoves or renames a file without touching its contents. -
old/path.txt deleteremoves a file outright.
That last one matters because the old behavior changed slightly: an empty REPLACE block used to mean "delete this file," but only if the SEARCH block matched the file's entire current contents. That meant deleting a file required the chatbot to regenerate the whole thing, verbatim, inside the SEARCH block, just to throw it away.
That's wasted tokens at best, and at worst a single mismatched character in a large file would make the whole delete fail to apply. Now the chatbot just writes old/path.txt delete and moves on, no regeneration needed! An empty REPLACE block on its own now just clears the matched text and keeps the file (even if that leaves it empty).
Undo now warns you before it steps on changes
If you hit Undo on an applied diff and the file in question has since been modified, either on disk or with unsaved changes in the editor, ReptClip now shows a warning first instead of silently reverting over your work.
And... I received my first external PR! 🥳
A huge part of making v2.0 safe to ship came from @mansio, who contributed two PRs to effessdev/reptclip, totalling roughly 1,500 lines of code!!! 🤯
The first PR focused on security in the diff-applying process. It uncovered several issues, including symlinks allowing writes outside the workspace, the ability to overwrite .git files, and Undo/Redo overwriting unsaved editor changes. It also fixed a .gitignore parsing bug.
The second PR caught even more edge cases, like missed editor changes on case-insensitive filesystems, race conditions during file updates, and incomplete symlink protection that could expose .git or .env contents to the chatbot. These are now fixed, huge thanks to @mansio! 🎉
A little help goes a long way 🥲
ReptClip wouldn't be quite what it is today without a few people who took the time to share their thoughts and suggestions.
A big shout-out to @unitbuilds for letting me know about the limited-time Qoder promotion. That tip made a huge difference and helped make ReptClip possible! The promotion may be over, but I now have something even better: ReptClip! 😁 In fact, I used ReptClip itself to build ReptClip v2. And I mean heavily! 🤯
Thanks to @piekwerk for actually giving ReptClip a try and suggesting that reverting changes should check for dirty editor tabs. I initially overlooked this, but now I get how important that was. In v2.0, both Undo and Redo now check for changes on disk and unsaved editor changes, warning you before proceeding.
And thanks to both @piekwerk and @mansio for pointing out something I hadn't considered: the order in which files are included in a prompt matters! Long prompts can lose effectiveness when important context gets buried in the middle. That feedback led me to make file patterns respect their written order, so you can push the most important files closer to the end of the context (right before your prompt).
For example:
-
**/*.py main.py: Includes all Python files, but putsmain.pylast. -
**/*.py src/**/*.py main.py: Includes all Python files, puts files undersrc/after the other Python files, and putsmain.pylast.
These might seem like small details, but they're exactly the kind of feedback that makes a project better. Thank you all for taking the time to help!
Try it out
Watch how beautifully ReptClip applies a large project reorganization:
Want to try it yourself? Head over to VS Code (or your favourite VS Code alternative), search for "ReptClip", and hit Install! It's that simple 😎
You can also find it on the official pages 🔥:
It's free, open source, and works with whatever chatbot you already have open in another tab. If you run into something weird while using it, please open an issue so I can fix it, or create a pull request if you want to fix it yourself 🙂 Reviews are fast right now! ⚡
If ReptClip saves you even one round of manual copy-pasting, or you think it'd be useful for others, the single most useful thing you can do for the project right now is head over to github.com/effessdev/reptclip and hit that ⭐ button! It's free, it takes two seconds, and it's the main way other people stumbling onto this post will know the project is worth a look!


Top comments (46)
Great👏
Thanks @xulingfeng bro!!!!! One word & emoji speak a lot 🫨
Wait wait wait...
You have starred 68 repositories and have NOT starred ReptClip yet??? Please star ReptClip... 🥺🙏 And also follow me (I followed you long ago bro). But star is more important so first star then follow (optional) but star is NOT optional... 🙏
lol me too same reaction I Also didn't understand what he meant
@xulingfeng What I meant is that you should go to github.com/effessdev/reptclip and click that ⭐ button ONCE.
Now get it?
yeah but why star is optional bro? she is famous 😆
I said star is NOT optional. Follow is optional, but star is mandatory.
bro but why xuling feng will not follow star? why it is optional? also he may have followed her
YEEEEEE!!!!! 🫨🫨🫨🫨🫨
bro I heard china tested 10g network why there is 5A written? or those 2 5A combined means 10A=10G?
Thanks @xulingfeng !!!!!
🤣
Bro don't make @xulingfeng unstar ReptClip lol
but what I did 😂 hehehehe
I'm a mind worm eating people's mind.
🧠🪱❓
Are you guys talking about me?😃
yeah I'm also figuring out
🤣
Piggybacking on the user's already-logged-in web session instead of forcing API keys is a super pragmatic pattern—the friction gap is almost never model intelligence anymore, it's context assembly in and deterministic diff application out.
Curious how forgiving your "Apply Diffs" parser is when a web chatbot silently normalizes indentation (tabs vs spaces or trailing whitespace) inside the
SEARCHblock? Line-trimmed fallback matching usually rescues ~90% of failed hunks without risking bad edits.Thanks for the comment! It's 4 layers. First, it tries with a strict check. If it cannot find a match, it tries to ignore line ending differences. If still cannot find a match, then it ignores trailing whitespaces, then shifted indentation, then finally tabs vs spaces. A unique match is still required at every level! And the status shows how many blocks were fuzzily matched, so you can press Undo and revert everything!!! Anything else you need to know or point out?!!
Requiring a unique match at every fallback level before applying is such a smart invariant - that alone prevents 90% of silent corruption when a file has repeated boilerplate blocks.
The only subtle multi-hunk edge case worth stress-testing is when an LLM emits two
SEARCH/REPLACEblocks for the same file and block #1 inserts or deletes lines above block #2:The .gitignore rules keep the junk out, but the files that bite are usually the tracked ones. A key that got committed once and never rotated will ride along happily when you hit Context, and a minute later it's sitting in some chatbot's conversation history on a server you don't control. Would it make sense to run a quick secret scan (or at least an extra ignore layer on top of the gitignore one) before the clipboard copy? Could sit next to showLineCounts as another toggle. Anyway, the mid-conversation command block rework is the part I'd actually reach for daily.
Thanks for the comment! Yeah, everything relies on the assumption that you have ignored all sensitive files. In most cases, accidentally committing a sensitive file is more serious than pasting it into a chatbot right? 🤔 So I thought any extra check wouldn't be worth it since it makes everything more complicated and less deterministic! 😌
Nice touch showing line counts in the tree so the model can choose between a full read and read-lines. One gap I'd watch: git-tracked doesn't mean text. Repos often track images, fonts or fixtures, and a broad glob like src/** will paste them as garbage. A cheap NUL-byte check (what git itself uses to flag binaries) would skip those. A rough token estimate shown before copying would also warn when the context won't fit the chatbot's window.
Thanks for your comment! 🙂 Good news: it already handles it!!! It shows something like "Binary file skipped" (I don't remember exactly 😔) instead of the actual contents! Binaries are detected using NUL-byte check like you said! It also has a 1MB limit. Meaning files over 1MB are also skipped, no matter whether they are binary or not! Both cases, no line count! If you attatch those files anyway, the chatbot can say "That file was skipped" since we aren't silently skipping it, but showing that it was skipped instead of the content! Files appear in the project structure normally, but with no line count! You can check the Safety measures for context section in the README for details!!! 🙂 Anything else you noticed? Let me fix them! 🔥
The git-tracked-only rule is the part that would actually get me to try this. Most clipboard-bridge tools make me nervous about what they'll touch, but limiting reads and writes to tracked files — and making a failed diff block the whole apply — that's the difference between a toy and something I'd run on a real repo. Curious how it handles a chatbot that returns the diff in a slightly different format though — does it ever just refuse and make you re-paste?
Thanks for the comment!!! If the diff is in a slightly different format, I will just refuse. We have to make sure the "AI instructions" checkbox is checked to include it in the context (which is checked by default!). We can also change the default configuration too! Using settings!
Bro!😊
Hi! 👋🙋♂️
Hi bro 😃
Bro can you like my this comment?😊
dev.to/star123/comment/3lh96
DONE!!!
Nice GIF! 😯
Thank you❤️
WHOOOOAOOAAAAA!!!!!! 🫨🫨🫨🫨🤯 Congrats!!!!! 👏👏👏👏👏
cool bro!😊
Thanks Star!!! 🌠
Bro, thank you so much for always being there and supporting me. Your kindness and support truly mean a lot to me. I’m genuinely grateful! ❤️
Hehe thanks a LOT star! 🙂🙂🫨😯🤯😳🙂 That comment looks like I wrote it lol 😮
All-or-nothing diff apply is the right call. One bad block should never leave a file half-written.
Thanks!!! I also think that feels more safe...
Great post
Isn't it? 😃
I Like it!😍❤️
Thanks DIDI!!!! 🙂