
一个强大的模型上下文协议服务器,用于LinkedIn交互,使AI助理能够搜索职位、生成简历和求职信,并以程序管理求职申请。
概览
一个MCP服务器,允许AI助理完全访问LinkedIn——搜索职位、查看个人资料和公司、生成AI驱动的简历和求职信,并跟踪申请。使用官方MCP Python SDK(FastMCP)构建。 —— 补充你的claude_desktop_config.json: 现代 ·专业 ·简易(简历)|专业 ·简明(求职信)HTML ·医学博士 ·PDF 有兴趣 ·申请 ·面试 ·录取 ·被拒 ·撤回 - — 使用 FastMCP 搭配 @mcp.tool() 装饰器,而非自定义协议实现 - — 所有同步LinkedIn API 调用都封装在 asyncio.to_thread() 中以避免阻塞 - — 工具→服务→客户端,服务层缓存 - — 核心LinkedIn功能无需 Anthropic API 密钥即可运行;AI 增强简历/求职信生成功能 - — Jinja2 SandboxedEnvironment、WeasyPrint SSRF 保护、路径穿越守卫、凭证编辑、输入验证 所有设置都是从环境变量(
概述(AI 翻译)
一个MCP服务器,允许AI助理完全访问LinkedIn——搜索职位、查看个人资料和公司、生成AI驱动的简历和求职信,并跟踪申请。使用官方MCP Python SDK(FastMCP)构建。
它的作用
| 类别 | 能力 |
|---|---|
| 求职 | 使用筛选功能搜索(关键词、地点、类型、经验水平、远程、近期情况),获取职位详情,获取推荐 |
| 个人资料与公司 | 获取任何LinkedIn个人资料或公司页面,AI驱动的个人资料分析及优化建议 |
| 简历生成 | 从LinkedIn个人资料生成简历,针对特定职位定制简历,内置3个模板 |
| 求职信 | AI生成的求职信,针对每个职位量身定制,内置2个模板 |
| 申请跟踪 | 本地跟踪申请,使用状态工作流程(感兴趣→申请→面试→录用/拒绝/撤回) |
| 输出格式 | HTML、Markdown 和 PDF(通过 WeasyPrint) |
快速启动
1.安装
2.配置
编辑 .env 并填写你的资历:
3.跑
独立:
使用 Claude Desktop — 添加到你的 claude_desktop_config.json:
本页由 AI 翻译,技术细节请以原文为准。
README
以下为原文 README 节选,完整内容请查看原文。
LinkedIn MCP Server
An MCP server that gives AI assistants full access to LinkedIn — search jobs, view profiles and companies, generate AI-powered resumes and cover letters, and track applications. Built with the official MCP Python SDK (FastMCP).
What It Does
| Category | Capabilities |
|---|---|
| Job Search | Search with filters (keywords, location, type, experience level, remote, recency), get job details, get recommendations |
| Profiles & Companies | Fetch any LinkedIn profile or company page, AI-powered profile analysis with optimization suggestions |
| Resume Generation | Generate resumes from LinkedIn profiles, tailor resumes to specific job postings, 3 built-in templates |
| Cover Letters | AI-generated cover letters personalized to each job, 2 built-in templates |
| Application Tracking | Track applications locally with status workflow (interested → applied → interviewing → offered/rejected/withdrawn) |
| Output Formats | HTML, Markdown, and PDF (via WeasyPrint) |
Quick Start
1. Install
# Core installation
pip install -e .
# With AI features (resume/cover letter generation, profile analysis)
pip install -e ".[ai]"
# With PDF export
pip install -e ".[pdf]"
# Everything
pip install -e ".[all]"
2. Configure
cp .env.example .env
Edit .env with your credentials:
[email protected]
LINKEDIN_PASSWORD=your_password
ANTHROPIC_API_KEY=sk-ant-... # Optional — enables AI features
3. Run
Standalone:
linkedin-mcp
With Claude Desktop — add to your claude_desktop_config.json:
{
"mcpServers": {
"linkedin": {
"command": "linkedin-mcp"
}
}
}
With Claude Code — add to .mcp.json:
{
"linkedin": {
"command": "linkedin-mcp"
}
}
Tools Reference
Job Tools (3)
| Tool | Parameters | Description |
|---|---|---|
search_jobs |
keywords, location, job_type, experience_level, remote, date_posted, page, count |
Search LinkedIn jobs with rich filters |
get_job_details |
job_id |
Get full description, skills, and metadata for a job posting |
get_recommended_jobs |
count |
Get personalized job recommendations |
Profile Tools (3)
| Tool | Parameters | Description |
|---|---|---|
get_profile |
profile_id |
Fetch a LinkedIn profile ("me" for your own) — experience, education, skills |
get_company |
company_id |
Get company info — description, size, headquarters, specialties |
安装
This server does not publish a one-line install command.
Open the repository installation guide 新手?先看 mcpskill.org 教程 →配置
{
"mcpServers": {
"linkedin": {
"command": "linkedin-mcp"
}
}
}