External Model or Provider Selection
- Category
- Excessive Agency
- Confidence
- 90% confidence
- Finding
The skill explicitly redirects Claude Code prompts to a non-default third-party provider, which can materially change where user inputs, code, and agent context are processed. Without a clear warning and consent-oriented language, users may assume they are using the usual backend and unknowingly disclose sensitive data to an external model provider.
- Content
export ANTHROPIC_API_KEY="sk-kimi-..."
claude # Interactive / 交互模式 claude --print "Your prompt" # One-shot / 单次执行
text Claude Code auto-appends `/v1/messages`. No other changes.
