DeepSeek HarnessHub
プラグインを投稿

xiaoyizhuang03-droid/dsh-lmstudio-router ↗★ 0

dsh-lmstudio-router

Local LM Studio model router for the dsh web GUI: routes unimportant but token-expensive requests (bulk text work, compaction, session titles) to a local LM Studio model over its OpenAI-compatible /v1 endpoint, auto-launches LM Studio when it is not running, and fits every local request into the model's context window (default 8192). Dependency-free host plugin — mounted via cordis.patch.yml + a profile node_modules entry, no dsh source changes.

AI 分析

核心用途是通过本地模型分流非核心任务以节省 API 额度。适合本地拥有一定算力（运行 LM Studio）且希望降低 API 账单开销的用户。

生産性
パッケージ
dsh-lmstudio-router
バージョン
0.1.3
ライセンス
MIT
最終更新
2026/08/24

インストール

$npx -p @deepseek-ai/dsh dsh plugin --profile web add github:xiaoyizhuang03-droid/dsh-lmstudio-router

ドキュメント

README 全文を読む ↗

dsh-lmstudio-router

Local LM Studio model router for the dsh web GUI. Routes unimportant but token-expensive requests (bulk text work, compaction, session titles) to a local LM Studio model over its OpenAI-compatible /v1 endpoint, auto-launches LM Studio when it is not running, and fits every local request into the model's context window (default 8192). Dependency-free host plugin.

See README.zh.md for the full Chinese documentation.

Install

dsh plugin --profile web add link:

Then add a config override in ~/.dsh/profiles/web/cordis.patch.yml:

- id: lmstudio-router
  name: 'dsh-lmstudio-router'
  config:
    baseUrl: http://127.0.0.1:1234/v1
    model: qwen3.5-9b
    contextWindow: 8192
    autoLaunch: true

Restart dsh web. Force routing with [local] / [remote] (or local: / remote:) at the start of a user message.

Test

node scripts/selftest.mjs

License

MIT