A more coder friendly visual studio AI coding partner approach, as opposed to mindlessly tasked robot.

[dstevens79]

Select Topic Area

Product Feedback

Copilot Feature Area

Visual Studio

Body

Spoiler : i suck at wording so i had the ai write this up, the intent is clear i think.
This would be a game changer for how we interact with copilot in visual studio.
This addresses the problem of seeing it go off the rails and being forced to stop it completely, as opposed to nudging it in the right direction with insight.

Summary

GitHub Copilot’s current request-response interaction model blocks user input during execution. When Copilot is running a multi-step task (searching files, planning edits, executing tools), developers can’t inject small corrections or context without canceling the whole run. This loses valuable insight, slows iteration, and increases user frustration.

Proposal: Dual Chat Interface (“Think + Do”)

Think: Non-Blocking Chat

• Always available for input, even while execution is ongoing
• Accepts quick hints, clarifications, and course corrections
• Messages are treated as inline hints, queued and applied safely

Do: Execution Mode

• Handles tool calls, file edits, build/test runs, multi-step operations
• Shows progress/status as it does today
• Incorporates Think hints without a full restart

Example Workflow

Current (blocking):

User: "Fix cloud distribution"
Copilot: [Searches 5 files for 30s...]
User: [Knows exact location but must wait or cancel]

Dual interface (non-blocking):

User (Do): "Fix cloud distribution"
Copilot (Do): [Searching for cloud logic...]

User (Think): "It's in GenerateCloudCover around line 920"
Copilot (Do): ✅ Adjusting → jumping to line 920

User (Think): "Also reduce temp dependency to 0.15"
Copilot (Do): ✅ Added to edit plan → included in the same commit

Implementation Options

• Split-Pane UI: Side-by-side or above below Think chat (talking) and Do chat (code change execution) panes
• Floating “Nudge” Input: Lightweight context injector visible during execution
• Message Queuing: Think messages queued and applied between tool calls/checkpoints

Benefits

• Non-Blocking Collaboration: Users stay engaged while Copilot executes
• Faster Iterations: Correct mistakes mid-execution without restarts
• Better Context: Real-time developer knowledge feeds execution
• Reduced Frustration: No more “I knew the answer but couldn’t tell you”
• Preserved Transparency: Do pane continues showing full execution trace

Analogues

• Cursor AI: Side chat while main window codes
• Jupyter: Edit other cells while one executes
• VS Code Terminal: Input while build runs
• Slack threads: Continue discussing while automation progresses

Evaluation

• UX Value: High — addresses the “soft interrupt” gap in AI pair programming
• Feasibility: Medium — primarily UI/orchestration; minimal or no model changes
• Risks/Complexities:
  • Context integrity and guardrails (ensure hints don’t derail plans)
  • Ordering/concurrency (apply hints at safe checkpoints)
  • Traceability (audit which hints affected which steps)

Hypothesis (Telemetry Signals)

• High cancel rates during multi-step runs
• Frequent resubmissions with added context after cancellations
• Idle wait periods where users already know the right line/file

Request

Please consider adding a dual-mode interaction system:

1. Background execution (“Do” channel)
2. Concurrent lightweight chat (“Think” channel)
3. Safe context injection between tool calls without full restart

This would elevate Copilot from a request-response assistant to a true, real-time pair programming partner.

[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. ⭐

[dstevens79]

this is a perfect example of where an implementation like this would be useful :

each step was me hard stopping it :
to deal with it failing, then it deciding to do somethign i didnt want done.

these could have been "talking nudges" that altered the actions being taken or considered.