MoonshotAI / MoonshotAI/kimi-code
Instructions "do following"
Nobody has claimed this yet.
- Dominant language
- TypeScript
- Stars
- 7.5k
- Forks
- 1.2k
- Avg merge
- 11h 53m
- Merged PRs (30d)
- 350
Description
What version of Kimi Code is running?
0.20.1
Which open platform/subscription were you using?
kimi-code
Which model were you using?
kimi-for-coding
What platform is your computer?
Linux 6.19.11-100.fc42.x86_64 x86_64 unknown
What issue are you seeing?
Completely and consistently ignores user instructions to "do this way" When reasoning, it either says, "It's too long/comprehencive, let's try this..." or "The user asked for..., but let's do... first." I have to persuade and prove my case, and sometimes this takes a lot of time or even doesn't help
What steps can reproduce the bug?
Have multistep task with 50+ equal but not the same activityes. Ask not to use any coding tools and review itself by reading each. This is just one case. I have other coding cases where I ask implement following feature, but have reply in thinking comments I described above.
What is the expected behavior?
follow instructions.
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 reproducing the report with a 50-plus-step task and the instruction not to use coding tools, then compare the agent's reasoning with the requested constraints. The payload names no source file, entry point, or test, so locating the instruction-following path and defining reliable passing criteria are part of the initial investigation.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- typescript
- Domain
- ai, cli
- Issue type
- Bug
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 30/100