Override kimi baseUrl, warming of exseed context limit

[bruce6135]

Description

Steps to reproduce

step1, The file ~\.codewhale\config.toml add following setting:
[providers.moonshot]
api_key = "sk-kimi-xxxxx"
base_url = "https://api.kimi.com/coding/v1"
step2, choose kimi-k2.7-code as current model, start a new conversasion, typing some words, the bottom shows warming msg: Emergency context compaction failed to reduce request below model limit (estimate ~21179 tokens, budget ~1024).

remark: the conversation is able to run in normal way, but every time I typing some words, it show the warming msg at bottom.

Expected behavior

Actual behavior

Impact

Environment

• OS: win10 x86/64
• codewhale version: 0.8.67
• Install method:
• codewhale doctor summary:

Version Information:
codewhale-tui: 0.8.67
rust: rustc 1.94.0 (4a4ef493e 2026-03-02)

Updates:
· current: v0.8.67
✓ latest: v0.8.67
Already up to date.

• Model/provider: kimi-k2.7-code / kimi coding plan
• Terminal app: win's 终端
• Shell:

<

Logs, screenshots, or recordings

[github-actions]

Thanks @bruce6135 for the report.

This issue is staying open for maintainer triage. CodeWhale gets better because people bring us real edge cases from real machines, providers, regions, and workflows.

If you can add a reproduction, logs, version output, screenshots, or the provider/model involved, that makes it much easier for us to verify and harvest the fix. Maintainers may comment /lgtmi to mark recurring issue reporters as approved so this intake note is skipped next time.

[mvanhorn]

Implemented in #4378.