[Critical Bug] Copilot Web (Claude Family) injects invalid whitespace into code syntax (PowerShell & Python)

[rayran]

Select Topic Area

Bug

Copilot Feature Area

Copilot in GitHub

Body

[Critical Bug] Copilot Web (Claude Family) injects invalid whitespace into code syntax (PowerShell & Python)

Severity: Critical (Generated code is syntactically invalid and fails execution immediately)
Platform: https://github.com/copilot (Web Interface)
Affected Models: Claude Sonnet (4 & 4.5), Claude Opus (4.5), Claude Haiku (4.5).

1. Executive Summary

I have confirmed through extensive A/B testing that ALL Claude-based models currently available on the Copilot Web UI are consistently generating invalid code. The models inject random whitespace characters (U+0020) immediately after the dot operator (.) in property accessors, method calls, and file extensions.

This behavior is systemic across languages (confirmed in both PowerShell and Python) and renders the generated code unusable without manual regex cleanup. Models from the GPT, Gemini, and Grok families do not exhibit this issue using identical prompts.

2. Technical Analysis & Error Patterns

The tokenization or decoding process for Claude models seems to be hallucinating spaces around punctuation marks.

🔴 PowerShell Failures

• Property Access: $obj. property instead of $obj.property
• Method Calls: $list. Add($item) instead of $list.Add($item)
• File Extensions: filename. json instead of filename.json
• Parameters: -Compress: $false (Syntactically risky/invalid)

🔴 Python Failures

• Attribute Access: self. pid, metric. to_dict() instead of self.pid, metric.to_dict()
• Module Calls: os. path. exists instead of os.path.exists
• Type Hinting: timestamp: datetime (Double spaces introduced)
• String Formatting: f"... {self. pid}..." (Spaces injected inside f-strings)

Note: The injected character is a standard ASCII space (U+0020),
not a zero-width or non-breaking space. This rules out encoding issues.

3. Evidence (Reproduction Proofs)

I have compiled "Shared Chat" links where I requested both PowerShell and Python scripts in the same session. The bug is present in every response from Claude models.

Status	Model	Shared Chat Evidence
❌ BROKEN	Claude Haiku 4.5	View Chat Log
❌ BROKEN	Claude Opus 4.5	View Chat Log
❌ BROKEN	Claude Sonnet 4	View Chat Log
❌ BROKEN	Claude Sonnet 4.5	View Chat Log

4. Control Group (Unaffected Models)

The following models generated syntactically correct code for the exact same prompts (no ghost spaces):

• Gemini Family: 2.5 Pro, 3 Flash, 3 Pro.
• GPT Family: 4o, 4.1, 5, 5.1, 5.2.
• Grok: Code Fast 1.

(Examples of successful generations available upon request, e.g., Gemini 2.5 Pro: https://github.com/copilot/share/8a195008-4140-8cf2-b003-2443646629c8)

5. Steps to Reproduce

1. Log in to https://github.com/copilot
2. Select Claude Sonnet 4 (or any Claude model).
3. Input the following prompt (or similar OOP request):

   "Act as a DevOps Engineer. Write a robust Python script using a 'SystemMetric' class to collect process data and export it to a JSON file."

4. Inspect the generated code block.
5. Notice the invalid spaces in self. attribute and json. dump.

6. Impact & Workaround

• Impact: Users must manually find-and-replace or use regex cleanup on every response
• Workaround: Switch to GPT or Gemini models until fixed
• Affected Users: Anyone using Claude models for code generation on Copilot Web

Haiku4.5.py
Opus4.5.py
Sonnet4.5.py
Sonnet4.py

[github-actions[bot]]

💬 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. ⭐

[invalidtask]

This has been a bug for at least two months. I do not understand why github hasn't addressed this glaring bug.