Bug type
Behavior bug (incorrect output/state without crash)
Summary
When receiving files via Feishu channel, file names containing Chinese characters are corrupted due to UTF-8 encoding being incorrectly interpreted as Latin-1/ISO-8859-1.
Steps to reproduce
- Configure Feishu channel in OpenClaw.
- Send a file with Chinese characters in the filename (e.g., "何不同舟渡.txt") via Feishu DM or group chat.
- Check the saved file path in OpenClaw's media/inbound directory.
Example:
- Original filename: "何不同舟渡_2.txt"
- Corrupted filename: "æµ_è_æ_ä_2---.txt"
Expected behavior
File names with Chinese characters should be preserved correctly. The saved file should have the original name "何不同舟渡_2.txt" or a sanitized version with Chinese characters intact.
Actual behavior
File names are corrupted. Chinese UTF-8 bytes are interpreted as Latin-1 characters, resulting in garbled filenames like "æµ_è_æ_ä_2.txt" instead of "何不同舟渡_2.txt".
This is a classic UTF-8 → Latin-1 mojibake where each Chinese character (3 UTF-8 bytes) becomes 3 separate Latin-1 characters.
OpenClaw version
2026.2.26
Operating system
Windows 11
Install method
npm global
Model
bailian/qwen3.5-plus
Provider / routing chain
openclaw -> bailian (Alibaba Cloud)
Config file / key location
No response
Additional provider/model setup details
No response
Logs, screenshots, and evidence
Impact and severity
Affected: Users receiving files with non-ASCII filenames via Feishu channel
Severity: Medium (files are saved correctly, but filenames are unreadable)
Frequency: 100% reproducible for files with Chinese/Japanese/Korean characters
Consequence: User inconvenience, difficulty identifying files, need to rename manually
Additional information
Workaround: Use English/ASCII filenames when sending files via Feishu, or rename files after receiving.
The issue likely occurs when OpenClaw processes the Content-Disposition header or file metadata from Feishu API, where UTF-8 encoded filenames are being decoded as Latin-1 instead of UTF-8.
Bug type
Behavior bug (incorrect output/state without crash)
Summary
When receiving files via Feishu channel, file names containing Chinese characters are corrupted due to UTF-8 encoding being incorrectly interpreted as Latin-1/ISO-8859-1.
Steps to reproduce
Example:
Expected behavior
File names with Chinese characters should be preserved correctly. The saved file should have the original name "何不同舟渡_2.txt" or a sanitized version with Chinese characters intact.
Actual behavior
File names are corrupted. Chinese UTF-8 bytes are interpreted as Latin-1 characters, resulting in garbled filenames like "æµ_è_æ_ä_2.txt" instead of "何不同舟渡_2.txt".
This is a classic UTF-8 → Latin-1 mojibake where each Chinese character (3 UTF-8 bytes) becomes 3 separate Latin-1 characters.
OpenClaw version
2026.2.26
Operating system
Windows 11
Install method
npm global
Model
bailian/qwen3.5-plus
Provider / routing chain
openclaw -> bailian (Alibaba Cloud)
Config file / key location
No response
Additional provider/model setup details
No response
Logs, screenshots, and evidence
Impact and severity
Affected: Users receiving files with non-ASCII filenames via Feishu channel
Severity: Medium (files are saved correctly, but filenames are unreadable)
Frequency: 100% reproducible for files with Chinese/Japanese/Korean characters
Consequence: User inconvenience, difficulty identifying files, need to rename manually
Additional information
Workaround: Use English/ASCII filenames when sending files via Feishu, or rename files after receiving.
The issue likely occurs when OpenClaw processes the Content-Disposition header or file metadata from Feishu API, where UTF-8 encoded filenames are being decoded as Latin-1 instead of UTF-8.