创建一个文件夹叫Project_Sys_Memory，放两个文件

Project_Preference.md:

# Definition
This file containes user preferences for developers & code agents for this project.

## About Git
- When init a project you should make sure you init the root path of the current repo as a git repo and go on with brach main (unless it is already one).
- I would inspect the code and commit manually. So you only need to append recommended commit command at the end of each final response, so I could edit and paste, then excute it when I think necessary. (Command example: `git commit -m "Added parallel request processing".)
- Don't forget to create and update .gitignore when updating the code.
- For files like test-run script (create the script only if i asked you to do so) or env files, make them ignored so when updating something they don't always diff. And, create .example files for them as template.
- ./Project_Sys_Memory path itself should be gitignored
- Project memory files like CLAUDE.md should be gitignored

## About Docker
- I prefer using docker primarily for project building. Each time code got updated I would manually excute `up --build -d` to test (or deploy), don't do that for me.
- For things that are not suitable for docker deployment, it is ok to go without docker.
- Code should not be directly linked into container, use copy when building, even this makes build needed everytime. I like going this way to avoid file related issues.
- Service inside docker should listen to 0.0.0.0 unless I specifically ask you to limit it from outer access
- If using docker compose, command should be `docker compose` not `docker-compose`, docker-compose was the engine v1 old way and not supported in v2 engine.

# About Env
- I like using .env files to configure projects. Create .env.example if needed.

# About Memory
- If you have project memory file, you should link all files under ./Project_Sys_Memory into the project memory file to keep them strictly followed through developing and not ignored after a long workflow.
- I would make a commmit when I confirm some code edited in the repo was working (expect those start with `[Experimental]`). So always check git's status and evaluate the project's change each time you start working (triggered by me), make it the 1st todo in the task list. If you found anything in project memory file needs to be edited/added/removed after the evaluation, you should do so before other work starts.
- You should only create or edit notes/memory files after I have made a commit. And you must check and update memory files after I have made a commit.

Project_Definition.md:

# Project Definition

# Your project definition