Repository navigation
Updated System Instructions to avoid bad state/conflict #17753
Description
Activity
- added a parent issue
on Jan 28, 2026 Hey @abhipatel12, I'd love to help with this issue if possible! I know it’s tagged for maintainers, but please let me know if you’d be open to me submitting a PR for it. Thanks!"
- added 3 commits that reference this issue
on Mar 5, 2026 - added a commit that references this issue
on Mar 6, 2026 - added a commit that references this issue
on Mar 12, 2026 - added a commit that references this issue
on Mar 12, 2026 - added a commit that references this issue
on Mar 14, 2026 - added a commit that references this issue
on Mar 21, 2026 - added a commit that references this issue
on Mar 30, 2026 - added a commit that references this issue
on Apr 9, 2026 - added a commit that references this issue
on May 6, 2026 - added a commit that references this issue
on Jul 9, 2026
The primary goal of this issue is to add some level of steering for the main agent to help steer the model from initializing subagents that will most likely have conflicts and cause a bad state.
This is not the end state solution but an attempt to rely on the model's reasoning to avoid obvious bad state. In this early version of parallel agents, users must also be thoughtful in how these agents are invoked. We may have more complex locking/worktree/etc solutions to solve this.
Criteria: