💬 Your Product Feedback Has Been Submitted 🎉

Thank you for taking the time to share your insights with us! Your feedback is invaluable as we build a better GitHub experience for all our users.

Here's what you can expect moving forward ⏩

• Your input will be carefully reviewed and cataloged by members of our product teams. 
  • Due to the high volume of submissions, we may not always be able to provide individual responses.
  • Rest assured, your feedback will help chart our course for product improvements.
• Other users may engage with your post, sharing their own perspectives or experiences. 
• GitHub staff may reach out for further clarification or insight. 
  • We may 'Answer' your discussion if there is a current solution, workaround, or roadmap/changelog post related to the feedback.

Where to look to see what's shipping 👀

• Read the Changelog for real-time updates on the latest GitHub features, enhancements, and calls for feedback.
• Explore our Product Roadmap, which details upcoming major releases and initiatives.

What you can do in the meantime 💻

• Upvote and comment on other user feedback Discussions that resonate with you.
• Add more information at any point! Useful details include: use cases, relevant labels, desired outcomes, and any accompanying screenshots.

As a member of the GitHub community, your participation is essential. While we can't promise that every suggestion will be implemented, we want to emphasize that your feedback is instrumental in guiding our decisions and priorities.

Thank you once again for your contribution to making GitHub even better! We're grateful for your ongoing support and collaboration in shaping the future of our platform. ⭐

I can confirm I just had this happen in 6 places in a single GitHub action workflow file I had Claude Opus 4.5 modify from the GitHub web page.

It took lines like this:

run: echo "${{ secrets.GITHUB_TOKEN }}" | docker login ghcr.io -u ${{ github.actor }} --password-stdin

And gave me back this, with no other changes to this line except the new space after ghcr.:

run: echo "${{ secrets.GITHUB_TOKEN }}" | docker login ghcr. io -u ${{ github.actor }} --password-stdin

This has been happening since Mon, I thought it is an issue with the editor:

Look at this diff

this is still there, even in chats:

This is incredibly frustrating. It happens on github in browser, in app, and in vscode with Claude Opus 4.5. It happened as recent as moments before I posted this reply.

this is incredibly frustrating and isn't limited to Claude. It is as though Copilot is treating periods as grammar vs code. I am having to manually check each piece of code with search and replace.

🐛 Fixing Space After Decimals in Claude Opus (Copilot) Output

This document addresses the issue where the Claude Opus model, when accessed through the GitHub Copilot interface, incorrectly inserts a space after the decimal point in floating-point numbers (e.g., rendering 1.000 as 1. 000).

1. 🔍 Diagnosis of the Problem

The issue is likely not a bug in the Claude Opus model itself, but rather a tokenization or rendering error introduced by the GitHub Copilot integration layer when processing Claude's output.

• Evidence: The core Claude.ai website does not reproduce this formatting error, and the model struggles to "see" and correct the error when asked within the Copilot environment.
• Impact: This defect forces manual cleanup of generated code, particularly for initialization values, mathematical equations, and configuration variables.

2. 🛠️ Workaround: Manual Correction

Since the issue is external to the model's logic, the immediate solution is manual cleanup using editor features.

A. Find and Replace

Use your IDE's global or file-specific Find and Replace feature.

B. Use Code Formatters

If you are generating configuration files or code in a language with strict formatting rules (e.g., Python, JavaScript, C++), running a dedicated code formatter (like Prettier, Black, or clang-format) immediately after generation may resolve the issue, as these tools prioritize strict syntax.

3. 📢 Providing Effective Feedback

As this is an integration/rendering bug, the most critical step is providing detailed, specific feedback to the GitHub Copilot team.

A. Report the Bug via GitHub/IDE

• Use the official feedback mechanism within your IDE (e.g., Visual Studio Code or Visual Studio's Copilot feedback tool).
• Key Details to Include:
  1. Model Used: Claude Opus 4.5.
  2. Environment: GitHub Copilot Chat / Interface.
  3. Specific Example: Provide the exact code prompt you used and the exact faulty output (e.g., Prompt: Write a variable with a default of 1.000. Output: var = 1. 000;).
  4. Crucial Observation: State that the bug is not reproducible on Claude.ai directly, suggesting the fault lies with the Copilot's handling of the output stream.

B. Mention the Tokens

• The issue may stem from Claude using a token for the decimal point and the Copilot environment improperly interpreting a following space token (which should be negligible) as a mandatory character, or a specific character encoding/token splitting issue between the services.

I am also experiencing this, and it's a big problem

Same problem.

It used to work fine, now feels like all generated code has a space after each dots, and the code is not usable. It is not an option for me to replace those manually each time.

Same problem. "Qwen2.5-VL-7B-Instruct. mmproj-F16.gguf"

hi there, i am getting this error on claude 4.5 sonnet and opus. there is no way to address the issue.
i found out if you create an agent in a repo that is existed before and has no issue, the agent works well and there is now mistaken whitespace. but the output code in the github copilot web view will put white space after "." in many cases. take a look on the screenshot that indicates this error : ( i encounter this problem just now !)

This has been going on in Copilot for the latest versions (4.5) of the top two models from Anthropic (Sonnet, Opus) since the day they were included in Copilot.

The only workaround I have found so far is to use OpenAI's cheapest coding model i.e. GPT-4.1 to "clean the redundant spaces, occurring mostly after punctuation marks (including in strings and comments)".

Please note that an affected model cannot be used for this cleanup - it sees the problem, attempts to fix it, but fails to do so.

Let's escalate this to Github Issues visited by Anthropic devs, who are the only people interested that these models continue to be used, (Microsoft's Copilot team has hedged their bets with all other major model providers:), like I did here:
anthropics/claude-code#10969

Feel free to do the same if you find anything remotely similar :)