MoonshotAI / MoonshotAI/kimi-cli
Tool call safety configuration/parameters
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 11.4k
- Forks
- 1.3k
- Avg merge
- 9h 47m
- Merged PRs (30d)
- 2
Description
What feature would you like to see?
It would be good to be able to configure which commands are allowed or disallowed in bash, and also to restrict reading/writing files to only a given directory. At the moment you can either go with afk/yolo which just allows everything, or you have to approve every single tool call including things like WriteFile and ReadFile and StrReplace within the project. For example allow listing/deny listing bash command patterns, or directories in which reading/writing/replacing is allowed. At the moment the only solution to guarantee security is to run in a container/VM or add guidelines and hope for the best, but it would be useful to have hard coded/deterministic guardrails built into the agent software that would prevent certain tool calls from being possible.
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
The issue names bash, ReadFile, WriteFile, and StrReplace as the affected tool calls, but no files or tests. Start by locating the approval and tool-execution entry points, then determine how command patterns and permitted directories should be configured and enforced. Done means disallowed commands and file operations are deterministically blocked without requiring per-call approval.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- cli, security
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100