MoonshotAI / MoonshotAI/kimi-cli
Plan mode: agent loops indefinitely on Bash echo / ReadFile instead of writing plan (kimi-code 0.38.0, K3)
Open
Nobody has claimed this yet.
bug
- Dominant language
- Python
- Stars
- 11.4k
- Forks
- 1.3k
- Avg merge
- 9h 47m
- Merged PRs (30d)
- 2
Description
What version of Kimi Code CLI is running?
0.38.0
Which open platform/subscription were you using?
uv
Which model were you using?
k3
What platform is your computer?
linux
What issue are you seeing?
Plan mode 下，探索完成后模型不执行写计划/ExitPlanMode，
而是死循环重复调用：
- Bash echo <任意字符>
- ReadFile 同一文件
每次人工打断并告知"不要循环"后，模型表面认错，
下一轮又进入同样的循环。
What steps can reproduce the bug?
- kimi --plan 进入 plan 模式
- 给它一个需要较多探索的任务
- 探索完成后观察行为
What is the expected behavior?
探索完成后应写入 plan 文件并调用 ExitPlanMode。
Additional information
kimi-session-session_af4f4c16-51c8-446d-9892-53c2406d2e86.zip
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
Reproduce the issue with kimi --plan using a task that requires substantial exploration, then inspect the attached session archive for the repeated Bash echo and ReadFile calls. Trace the plan-mode flow and identify why it does not write the plan or call ExitPlanMode. Done means exploration terminates with a plan file and a single ExitPlanMode call instead of looping.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- ai, cli
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Active
- Clarity
- Mostly clear
- Newbie friendliness
- 52/100