Claude Code Token Saver MCP Server

@csabakecskemeti

代码仓库

基本信息

分类

开发工具

传输方式

stdio

作者

csabakecskemeti

提交者

Csaba Kecskemeti

关于 Claude Code Token Saver MCP Server

The MCP server allows users to utilize user-defined and/or local LLMs in Clause Code to handle smaller, straightforward tasks and save tokens.

Claude Code Token Saver MCP Server 配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "cc-token-saver": {
      "type": "stdio",
      "command": "python",
      "args": [
        "<path>/cc_token_saver_mcp/server.py"
      ]
    }
  }
}

Claude Code Token Saver MCP Server 工具

工具列表暂未提供

当前收录暂时没有可展示的工具列表，不代表服务没有工具。请查阅项目文档了解支持的能力。

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is Claude Code Token Saver?

Claude Code Token Saver is an MCP server that reduces token usage in Claude Code by delegating simple tasks—like code snippet generation, documentation, and basic Q&A—to a local LLM, reserving Claude Code for complex coordination and architecture decisions.

How to use Claude Code Token Saver?

Set up the MCP server and configure Claude Code to use it. Claude Code automatically tries the local LLM first for simple tasks; you invoke the server’s tools (e.g., query_local_llm) through standard MCP tool calls. No special commands beyond normal MCP configuration are required.

Key features of Claude Code Token Saver

Delegates simple tasks to your local LLM automatically.
Exposes local LLM as MCP tools for Claude Code.
Saves premium tokens for complex reasoning workflows.
Supports code generation, refactoring, documentation, and reviews.
Works with any local LLM compatible with the MCP server.

Use cases of Claude Code Token Saver

Generating boilerplate or example code without using premium tokens.
Performing quick code reviews and refactoring tasks locally.
Writing documentation and answering basic programming questions.
Offloading simple, repetitive coding tasks to conserve token budgets.
Keeping Claude Code focused on multi-step architecture decisions.

常见问题

What tasks does the local LLM handle?

Simple tasks such as code snippet generation, basic refactoring, documentation writing, code reviews, and basic Q&A.

How does Claude Code Token Saver decide which tasks to delegate?

Claude Code automatically tries the local LLM first for simple tasks; only complex reasoning and multi-step workflows are sent to Claude Code’s premium tokens.

Do I need to run a local LLM to use this server?

Yes, the server requires your local LLM to be accessible; it exposes that local LLM as tools for Claude Code.

What configuration is needed?

No specific configuration details are provided in the README beyond standard MCP server setup and ensuring your local LLM is available.

Are there any known limitations?

— (No limitations mentioned in the README.)

开发工具分类下的更多 MCP 服务器

Routebase

Routebase keeps your API's design, docs, contract tests, mocks and monitoring in one living OpenAPI spec so they can't drift apart. Agents work on that same source of truth, under your team's permissions.

DocBin

Give an agent a place to put large output. Instead of returning 40k characters of a build log into the conversation, it posts the file to DocBin and hands back one short URL. Docs render with syntax highlighting and mark

LastPing

Know when your AI agents, cron jobs and CI pipelines fail, stall or go quiet. Your agent sets up monitors from a plain description, tells you what is broken and why, reads run history and traces, and writes notes on inci

Capgo MCP

Cap-go

Console, Backend and CLI to manage Capgo Instant update and Native build for Capacitor apps Build and upload to app store your mobile app

Casefile

azimov777

One case file per task — shared by every session, agent and person that touches it. Self-hosted MCP server and live board. MIT.