MoonshotAI / MoonshotAI/kimi-code
Feature request: option to disable automatic EnterPlanMode
Nobody has claimed this yet.
- Dominant language
- TypeScript
- Stars
- 7.5k
- Forks
- 1.2k
- Avg merge
- 11h 53m
- Merged PRs (30d)
- 350
Description
What feature would you like to see?
The agent sometimes calls the EnterPlanMode tool on its own initiative — without me ever requesting Plan Mode. For example:
● I'll plan this out first — it's a multi-file setup with a few toolchain decisions to make.
● Used EnterPlanMode
This unilaterally switches my session into Plan Mode, which:
- Interrupts the flow. I asked for the task to be done, not for a plan review. Now I have to approve/reject a plan before anything happens — an extra round-trip I never opted into.
- Overrides my permission-mode choice. If I wanted planning, I would toggle Plan Mode myself. The mode switch is a user-level decision; the model shouldn't make it for me.
- Is unpredictable. Whether a task triggers auto-planning depends on the model's judgment of "non-trivial", so identical workflows behave differently run to run.
Proposed solution
At minimum, provide a config option to disable the agent-initiated EnterPlanMode, e.g. in config.toml:
# Never let the model enter Plan Mode on its own;
# Plan Mode is only entered when the user toggles it explicitly.
auto_plan_mode = false
Additional information
No response
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by locating the configuration handling for config.toml and the EnterPlanMode tool invocation. Determine how to distinguish agent-initiated entry from an explicit user toggle, then verify that auto_plan_mode = false prevents only the automatic path while preserving manual Plan Mode entry.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- typescript
- Domain
- cli, tooling
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Quiet
- Clarity
- Mostly clear
- Newbie friendliness
- 45/100