Code Refactor MCP Server

by zhouhengchangio.github.zhouhengchang/code-refactorv1.0.4

Refactor legacy code into modern, idiomatic code without changing its behavior.

0Node.jsstdioremoteofficial registry

context tax

~297 tok/turn2 tools

security

97/100

cold start

pending

freshness

Active1d ago

Our reading of Code Refactor, from the data

Code Refactor is an MCP server from zhouhengchang, listed in the official MCP Registry.

Its 2 tools add about 297 tokens to every model turn (measured 2026-10-19).

Maintenance is active: the last public signal was in the last day (2026-10-18).

It runs either locally over stdio or as a hosted endpoint and needs one secret (OPENAI_API_KEY); use a scoped, least-privilege key.

Install Code Refactor MCP server

claude mcp add zhouhengchang -e OPENAI_API_KEY='<openai-api-key>' -- npx -y workbuddy-code-refactor-connector

{
  "mcpServers": {
    "zhouhengchang": {
      "command": "npx",
      "args": [
        "-y",
        "workbuddy-code-refactor-connector"
      ],
      "env": {
        "OPENAI_API_KEY": "<openai-api-key>"
      }
    }
  }
}

Add to ~/.cursor/mcp.json (global) or .cursor/mcp.json (project).

{
  "mcpServers": {
    "zhouhengchang": {
      "command": "npx",
      "args": [
        "-y",
        "workbuddy-code-refactor-connector"
      ],
      "env": {
        "OPENAI_API_KEY": "<openai-api-key>"
      }
    }
  }
}

Settings → Developer → Edit Config (claude_desktop_config.json), then restart.

{
  "servers": {
    "zhouhengchang": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "workbuddy-code-refactor-connector"
      ],
      "env": {
        "OPENAI_API_KEY": "<openai-api-key>"
      }
    }
  }
}

Add to .vscode/mcp.json in your workspace.

{
  "mcpServers": {
    "zhouhengchang": {
      "command": "npx",
      "args": [
        "-y",
        "workbuddy-code-refactor-connector"
      ],
      "env": {
        "OPENAI_API_KEY": "<openai-api-key>"
      }
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json.

Configuration

Variable	Required	Secret	Description
REFACTOR_BACKEND	—	—	Inference backend: 'ollama' (local, code stays on your machine, default) or 'openai' (any OpenAI-compatible cloud endpoint, e.g. DeepSeek).
OPENAI_BASE_URL	—	—	Base URL of the OpenAI-compatible endpoint, e.g. https://api.deepseek.com/v1. Required when REFACTOR_BACKEND=openai.
OPENAI_API_KEY	—	yes	API key for the cloud endpoint, e.g. a DeepSeek key. Only needed when REFACTOR_BACKEND=openai.
OPENAI_MODEL	—	—	Cloud model name, e.g. deepseek-chat.
OLLAMA_HOST	—	—	Ollama server address used when REFACTOR_BACKEND=ollama.
OLLAMA_MODEL	—	—	Ollama model name used when REFACTOR_BACKEND=ollama.
REFACTOR_TIMEOUT_MS	—	—	Per-call timeout in milliseconds. Raise it for slow local CPU inference.

Remote endpoints

streamable-httphttps://code-refactor-api.app.workbuddy.host/mcp

Context tax

Connecting this server adds ~297 tokens of tool definitions to every model turn (lean, 2 tools). Measured 2026-10-19 via its hosted endpoint; o200k tokenizer, worst case for clients that load all tools up front. How we measure

Model	Per 1,000 turns	With prompt caching
Claude Opus 5.5	$1.19	$0.059
Claude Sonnet 5.5	$0.594	$0.059
Claude Haiku 5.5	$0.030	—

Input prices as of 2026-10-06. Excludes your own prompt and output.

Tools (2)

refactor_code184 tok
把遗留代码重构为现代规范代码，行为保持不变。会做语义化重命名、补全类型标注、生成中文注释、优化结构。适用于 Python / JavaScript / TypeScript / Java / Go / Rust / C / C++。
check_status63 tok
检查重构服务与云端模型（DeepSeek）是否可用，返回后端类型、模型名与密钥校验状态。建议在调用 refactor_code 前先调用本工具。

Security signals

97/100 — no known issues found as of 2026-10-19, from the checks below. This is not a guarantee of safety; review what a server can do before connecting it.

Deductions

No security policy (SECURITY.md)-3

Checks

✓Repository hygiene (archived, activity, license, security policy)
○Package metadata (install scripts, provenance) — pending
○Tool descriptions (prompt-injection and hidden-text heuristics) — pending
○Source scan (secrets, CVEs, risky code) by vectra-guard — pending

Scan it yourself before installing with vectra-guard, the open-source agent security CLI. Scoring rules

Freshness

Active — last maintenance signal 1d ago. The newest of the signals below sets the band.

Last commit (default branch)
2026-10-18 · 1d ago · GitHub
Latest release
no data · GitHub
Package published
no data · npm/PyPI
Registry entry updated
2026-10-18 · 1d ago · official registry · v1.0.4

FAQ

›How do I install the Code Refactor MCP server in Claude Code?

Run: claude mcp add zhouhengchang -e OPENAI_API_KEY='<openai-api-key>' -- npx -y workbuddy-code-refactor-connector. For Cursor, VS Code, Claude Desktop and Windsurf, use the install tabs above.

›Does Code Refactor require an API key?

Yes. It expects OPENAI_API_KEY, of which 1 is a secret.

›Can I use Code Refactor as a remote (hosted) MCP server?

Yes — it offers both a hosted endpoint and a local stdio package.

›Is Code Refactor in the official MCP registry?

Yes, as io.github.zhouhengchang/code-refactor.

›How many tokens does Code Refactor add to my context?

Its tool definitions total about 297 tokens, resent on every turn.

Sources and method

Official MCP Registry entry: updated 2026-10-18
GitHub repository: fetched 2026-10-19
Package registry (npm / PyPI): listed
Tool list (tools/list), measured by us: 2026-10-19
Security signals, computed by us: 2026-10-19
Last checked against all sources: 2026-10-19

The description, packages and links come from the maintainer's own published metadata. Token cost, security signals, freshness and the reading above are MCPSkillHub's own measurements; see the methodology. Spot an error? Request a correction or claim the listing.

Our reading of Code Refactor, from the data

Install Code Refactor MCP server

Install in Claude Code

Install in Cursor

Install in Claude Desktop

Install in VS Code

Install in Windsurf