[Bug]: extension.yml fails to read UTF-8 encoded Chinese content on Windows

[WANGMAN-Echo]

Bug Description

On Windows systems, when extension.yml contains UTF-8 encoded Chinese (or other non-ASCII) characters, the specify extension add command fails with the following error:
UnicodeDecodeError: 'gbk' codec can't decode byte 0xa5 in position 138: illegal multibyte sequence

Steps to Reproduce

1. Create an extension.yml with Chinese content:
   schema_version: "1.0"
   extension:
   id: "test"
   name: "测试扩展"
   version: "1.0.0"
   description: "这是一个测试扩展"
2. On Windows, run:
   specify extension add test --dev --from ./test-extension
3. Error occurs: UnicodeDecodeError: 'gbk' codec can't decode...
   Environment:

• OS: Windows 10/11
• Python: 3.x
• specify-cli: v0.5.0

Expected Behavior

The specify extension add command should correctly read and parse extension.yml files containing UTF-8 encoded Chinese or other non-ASCII characters, regardless of the operating system's default encoding.

Actual Behavior

The command fails with the following error:
UnicodeDecodeError: 'gbk' codec can't decode byte 0xa5 in position 138: illegal multibyte sequence

Specify CLI Version

v0.5.0

AI Agent

opencode

Operating System

Windows 11

Python Version

Python 3.14

Error Logs

Additional Context

No response

[mnriem]

Can you please validate if this is still an issue against the latest release?