Pre-submission checklist | 提交前检查
Bug Description | 问题描述
Bug Description
memos.exe (Rust binary) corrupts all Chinese/Japanese/Korean text on Windows. Characters are double-encoded: the binary internally converts UTF-8 input to the Windows system codepage (GBK/CP936) before sending to the API, causing irreversible data corruption.
How to Reproduce | 如何重现
Steps to Reproduce
# 1. Add a Chinese memory
memos add -m "测试中文记忆存储"
# 2. Search for it
memos search "中文记忆" --format json --detail simple
### Environment | 环境信息
### Environment
- **OS:** Windows 11 Enterprise 10.0.22631
- **CLI version:** `@memtensor/memos-cloud-cli@1.0.2` (Rust binary, PE32+ x86-64)
- **Shell:** Git Bash (MSYS2) — passes UTF-8 args correctly
- **System codepage:** ACP=936 (GBK), OEMCP=936 (GBK)
- **Node.js wrapper:** spawns `memos.exe` via `child_process.spawn` with `stdio: "inherit"`
### Additional Context | 其他信息
_No response_
### Willingness to Implement | 实现意愿
- [ ] I'm willing to implement this myself | 我愿意自己解决
- [ ] I would like someone else to implement this | 我希望其他人来解决
Pre-submission checklist | 提交前检查
Bug Description | 问题描述
Bug Description
memos.exe(Rust binary) corrupts all Chinese/Japanese/Korean text on Windows. Characters are double-encoded: the binary internally converts UTF-8 input to the Windows system codepage (GBK/CP936) before sending to the API, causing irreversible data corruption.How to Reproduce | 如何重现
Steps to Reproduce