Serena MCP服务器设置指南 - 语义代码分析与AI编程助手配置教程

mcp%3Asetup-serena-mcp by neolabhq/context-engineering-kit

219 周安装量

699 GitHub Stars

GitHub

安装命令

npx skills add https://github.com/neolabhq/context-engineering-kit --skill mcp:setup-serena-mcp

AI/机器学习开发自动化

🇨🇳中文介绍

$ARGUMENTS

Serena MCP 服务器设置指南

1. 确定设置上下文

询问用户希望将配置存储在何处：

选项：

项目级别（通过 git 共享） - 配置受版本控制跟踪，与团队共享
- CLAUDE.md 更新位置：./CLAUDE.md
项目级别（个人偏好） - 配置保留在本地，不受 git 跟踪
- CLAUDE.md 更新位置：./CLAUDE.local.md
- 验证这些文件是否已列入 .gitignore，若未列入则添加
用户级别（全局） - 配置适用于此用户的所有项目
- CLAUDE.md 更新位置：~/.claude/CLAUDE.md

存储用户的选择，并在后续步骤中使用相应的路径。

2. 检查 Serena MCP 服务器是否已设置

尝试使用 Serena MCP 服务器的某个工具（例如 find_symbol 或）来检查您是否有权访问它。

广告位招租

在这里展示您的产品或服务

触达数万 AI 开发者，精准高效

联系我们

3. 加载 Serena 文档

阅读以下文档以了解 Serena 的功能和设置过程：

加载 https://raw.githubusercontent.com/oraios/serena/refs/heads/main/README.md 以了解 Serena 是什么及其功能
加载 https://oraios.github.io/serena/02-usage/020_running.html 以学习如何运行 Serena
加载 https://oraios.github.io/serena/02-usage/030_clients.html 以学习如何配置您的 MCP 客户端
加载 https://oraios.github.io/serena/02-usage/040_workflow.html 以学习如何为您的项目设置 Serena

4. 引导用户完成设置过程

根据加载的文档：

检查先决条件：验证是否已安装 uv（运行 Serena 所需）
识别客户端类型：确定用户正在使用哪个 MCP 客户端（Claude Code、Claude Desktop、Cursor、VSCode 等）
提供设置说明：如果尚未配置，则引导完成针对其客户端的特定配置
设置项目：如果尚未设置，则引导完成项目设置过程
开始索引项目：如果刚刚设置完成，则引导完成项目索引过程
如果 MCP 刚刚设置完成，请要求用户重启 Claude Code 以加载新的 MCP 服务器，向用户明确写出说明，包括"退出 claude code 控制台，然后运行 'claude --continue'，然后输入 'continue' 以继续设置过程"
测试连接：验证设置后是否可以访问 Serena 工具
1. 如果还不行，运行 initial_instructions
2. 检查是否已执行 onboarding，如果没有则运行它。
3. 然后尝试读取任何文件

在添加 MCP 服务器之后，但在测试连接之前，请向用户精确地写入此消息：

You must restart Claude Code to load the new MCP server:

  1. Exit Claude Code console (type exit or press Ctrl+C)
  2. Run claude --continue
  3. Type "continue" to resume setup

  After restart, I will:
  - Verify Serena tools are accessible
  - Run initial_instructions if needed
  - Perform onboarding for this project (if not already done)

5. 更新 CLAUDE.md 文件

使用步骤 1 中确定的路径。一旦 Serena 成功设置，使用以下内容精确地更新相应的 CLAUDE.md 文件：

### 使用 Serena MCP 进行语义代码分析，而非常规代码搜索和编辑

Serena MCP 可用于高级代码检索和编辑功能。

**何时使用 Serena：**
- 基于符号的代码导航（查找定义、引用、实现）
- 在结构化代码库中进行精确的代码操作
- 在可用时，优先使用基于符号的操作而非基于文件的 grep/sed

**关键工具：**
- `find_symbol` - 在代码库中按名称查找符号
- `find_referencing_symbols` - 查找引用给定符号的所有符号
- `get_symbols_overview` - 获取文件中顶级符号的概览
- `read_file` - 读取项目目录内的文件内容

**使用说明：**
- 内存文件可以在 `.serena/memories/` 中手动查看/编辑

如果服务器设置在用户级别（全局），则添加此部分：

**项目设置（每个项目）：**
1. 在您的项目目录中运行 `serena project create --index`
2. Serena 自动检测语言；创建 `.serena/project.yml`
3. 首次使用会触发 onboarding 并在 `.serena/memories/` 中创建内存文件

6. 项目初始化（如果需要）

如果这是一个新项目或 Serena 尚未初始化：

引导用户运行项目初始化命令
解释基于项目的工作流程和索引
如果需要，配置项目特定的设置

🇺🇸English

User Input:

$ARGUMENTS

Guide for setup Serena MCP server

1. Determine setup context

Ask the user where they want to store the configuration:

Options:

Project level (shared via git) - Configuration tracked in version control, shared with team
- CLAUDE.md updates go to: ./CLAUDE.md
Project level (personal preferences) - Configuration stays local, not tracked in git
- CLAUDE.md updates go to: ./CLAUDE.local.md
- Verify these files are listed in .gitignore, add them if not
User level (global) - Configuration applies to all projects for this user
- CLAUDE.md updates go to: ~/.claude/CLAUDE.md

Store the user's choice and use the appropriate paths in subsequent steps.

2. Check if Serena MCP server is already setup

Check whether you have access to Serena MCP server by attempting to use one of its tools (e.g., find_symbol or get_symbols_overview).

If no access, proceed with setup.

3. Load Serena documentation

Read the following documentation to understand Serena's capabilities and setup process:

Load https://raw.githubusercontent.com/oraios/serena/refs/heads/main/README.md to understand what Serena is and its capabilities
Load https://oraios.github.io/serena/02-usage/020_running.html to learn how to run Serena
Load https://oraios.github.io/serena/02-usage/030_clients.html to learn how to configure your MCP client
Load https://oraios.github.io/serena/02-usage/040_workflow.html to learn how to setup Serena for your project

4. Guide user through setup process

Based on the loaded documentation:

Check prerequisites : Verify that uv is installed (required for running Serena)
Identify client type : Determine which MCP client the user is using (Claude Code, Claude Desktop, Cursor, VSCode, etc.)
Provide setup instructions : Guide through the configuration specific to their client if it not already configured
Setup project : Guide through the project setup process if it not already setup
Start indexing project : Guide through the project indexing process if it was just setup
If MCP was just setup, ask user to restart Claude Code to load the new MCP server, write to user explisit instructions, including "exit claude code console, then run 'claude --continue' and then write "continue" to continue setup process"
Test connection : Verify that Serena tools are accessible after setup 1. If not yet, run initial_instructions 2. Check if onboarding was performered, if not then run it. 3. Then try to read any file

After adding MCP server, but before testings connection write to user this message EXACTLY:

You must restart Claude Code to load the new MCP server:

  1. Exit Claude Code console (type exit or press Ctrl+C)
  2. Run claude --continue
  3. Type "continue" to resume setup

  After restart, I will:
  - Verify Serena tools are accessible
  - Run initial_instructions if needed
  - Perform onboarding for this project (if not already done)

5. Update CLAUDE.md file

Use the path determined in step 1. Once Serena is successfully set up, update the appropriate CLAUDE.md file with the following content EXACTLY:

### Use Serena MCP for Semantic Code Analysis instead of regular code search and editing

Serena MCP is available for advanced code retrieval and editing capabilities.

**When to use Serena:**
- Symbol-based code navigation (find definitions, references, implementations)
- Precise code manipulation in structured codebases
- Prefer symbol-based operations over file-based grep/sed when available

**Key tools:**
- `find_symbol` - Find symbol by name across the codebase
- `find_referencing_symbols` - Find all symbols that reference a given symbol
- `get_symbols_overview` - Get overview of top-level symbols in a file
- `read_file` - Read file content within the project directory

**Usage notes:**
- Memory files can be manually reviewed/edited in `.serena/memories/`

Add this section, if server setup at user level (global):

**Project setup (per project):**
1. Run `serena project create --index` in your project directory
2. Serena auto-detects language; creates `.serena/project.yml`
3. First use triggers onboarding and creates memory files in `.serena/memories/`

6. Project initialization (if needed)

If this is a new project or Serena hasn't been initialized:

Guide user to run project initialization commands
Explain project-based workflow and indexing
Configure project-specific settings if needed

Weekly Installs

219

Repository

neolabhq/contex…ring-kit

GitHub Stars

699

First Seen

Feb 19, 2026

Installed on

opencode212

codex211

github-copilot210

gemini-cli209

kimi-cli207

amp207

agent-browser 浏览器自动化工具 - Vercel Labs 命令行网页操作与测试

144,300 周安装