mcpField
C
Claude Code Token Saver

Claude Code Token Saver

Local
github.com/csabakecskemetiDeveloper Tools20Updated 71w ago
Repository

About Claude Code Token Saver

Claude Code Token Saver is an MCP server for Claude Code users who want a local LLM to handle straightforward coding and writing tasks. It exposes the local model as tools for smaller jobs, leaving complex reasoning and coordination to Claude Code.

Key features

  1. 01Delegate code snippets to a local LLM
  2. 02Handle simple refactoring tasks
  3. 03Draft documentation
  4. 04Review code
  5. 05Answer basic questions

Use cases

  1. 01Generate a small code snippet locally
  2. 02Apply a straightforward code refactor
  3. 03Draft or review project documentation
  4. 04Run a basic code review or Q&A

MCP support

Exposes a configured local LLM to Claude Code through MCP tools, which Claude Code can use for smaller tasks.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

No install method listed yet

We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.

Repository

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

cc_token_saver_mcp

Allow Claude code to use local llm for smaller tasks to save token or for specialized task.

Reduce your Claude Code tokens with ‘CC token saver’ MCP server that intelligently delegates simple tasks to your local LLM while keeping Claude Code for complex coordination and architecture decisions.

The MCP server exposes your local LLM as tools that Claude Code can use for:

  • Code snippet generation
  • Simple refactoring tasks
  • Documentation writing
  • Code reviews
  • Basic Q&A Claude Code automatically tries the local LLM first for simple tasks, only using premium tokens when necessary for complex reasoning and multi-step workflows.

MCP server config

Create a .env file with the LLM config

Example:

# Local LLM Configuration
OPENAI_API_KEY=none
OPENAI_BASE_URL=http://localhost:1234/v1
LOCAL_MODEL_NAME=qwen2.5-7b-instruct
LOCAL_LLM_TEMPERATURE=0.7
LOCAL_LLM_MAX_TOKENS=-1

Claude Code MCP config

edit the ~/.claude.json file

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

Example usage: cc-token_saver

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.