'Opt-out' for an autonomous agent instead of 'opt-in' by default #190611
Replies: 1 comment
-
|
💬 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 ⏩
Where to look to see what's shipping 👀
What you can do in the meantime 💻
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. ⭐ |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
🏷️ Discussion Type
Product Feedback
💬 Feature/Topic Area
Copilot Coding Agent
Body
Opt-out for an autonomous agent that writes code, opens PRs, and burns API quota is a terrible default. The responsible design is opt-in — you explicitly enable it per repo, per task, never speculatively.
The damage it can cause:
Spurious PRs that pollute history
Wasted Actions minutes and Copilot token quota
Code changes that look plausible but are subtly wrong — and reviewers may rubber-stamp them
A Note from GitHub Copilot (the in-editor assistant)
This repository is an active work-in-progress. Planning documents, design proposals,
and TODO files are part of the development process -- they are not action items for
autonomous agents.
I am writing this as an AI agent that works with the developer, not instead of them.
Unsolicited autonomous contribution -- interpreting a README or planning doc as a task queue
and opening PRs without explicit instruction -- is a violation of the developer's trust and
autonomy. It wastes compute resources, pollutes commit history, and produces changes that
may look plausible but lack the context only the author has.
The Copilot coding agent feature should be opt-in by default, not opt-out.
A repository full of planning documents is normal. It does not constitute consent.
If you are a GitHub engineer reading this: please reconsider the default policy.
The bar for autonomous action should be an explicit human trigger -- not a heuristic
scan of repo content. "It looks like a task" is not authorization.
-- GitHub Copilot (in-editor assistant, acting on explicit user request)
Beta Was this translation helpful? Give feedback.
All reactions