linkedin-mcp
linkedin-mcp
用于从 Claude Code 编辑你自己的 LinkedIn 个人资料的 MCP 服务器——采用硬性的 propose-then-approve(先提议后批准) 分离机制:模型可以起草变更并向你展示 diff,但整套流程中只有一个工具能执行写入,而且它只会针对你已经查看过的提议执行。
合作伙伴准入门槛: LinkedIn 的 Profile Edit API 仅对 LinkedIn 批准的合作伙伴开发者开放(
w_compliance是“一项私有权限,仅授予选定的开发者”——参见 docs/api-notes.md)。在合作伙伴批准落地之前,apply_proposal会返回 invalid-scope(作用域无效)错误。这是预期行为,不是缺陷。 其他一切(OAuth、提议、diff)目前都正常工作。
工作方式
工具 | 说明 | 是否会写入 LinkedIn? |
| 输出 LinkedIn OAuth URL,捕获一次性 localhost 重定向,换取 code,存储令牌 | 否(仅 OAuth) |
| 报告令牌是否存在以及何时过期 | 否 |
| 获取你的个人资料( | 否 |
| 为某项变更构建确切的 API 请求(headline、summary、positions、skills、educations),将其保存为提议,并返回 unified diff + | 绝不 |
| 列出已保存的提议 | 否 |
| 删除已保存的提议 | 否 |
| 向 LinkedIn 发送一个已保存的提议—— 唯一的写入工具。由代码强制执行的确认门禁:只有当调用时传入了 | 是 |
提议会持久化保存在 ~/.config/linkedin-mcp/proposals/ 下,因此批准可以在之后的会话中完成。令牌和客户端凭据存储在 ~/.config/linkedin-mcp/.env 中,权限为 600——由你亲自填写,绝不会由 agent 写入,也永不提交(参见 .env.example)。
Related MCP server: kaushik-linkedin-mcp
安装
需要 Python ≥ 3.11。
{
"mcpServers": {
"linkedin": {
"command": "bash",
"args": ["/path/to/linkedin-mcp/run.sh"]
}
}
}run.sh 在首次启动时会创建 .venv/ 并安装锁定依赖(以 stamp 文件为门禁;所有启动引导输出都会写入 stderr,从而保持 MCP stdio 通道干净)。
设置
按 docs/SETUP-KYLE.md 操作——它会引导你创建 LinkedIn Developer 应用、注册 redirect URL、填写 ~/.config/linkedin-mcp/.env、运行 auth_start,并申请合作伙伴计划。
开发
bash run.sh --help # bootstraps the venv (runtime deps only)
.venv/bin/pip install -r requirements-dev.txt # adds pytest + coverage tooling
.venv/bin/python -m pytest # offline — every test runs against a mock transport测试套件中包含一个 granted-write fixture:一个 mock LinkedIn,其中每个文档中写端点都会愉快地返回 200。测试断言:即使写入可能成功,propose_edit 也会在记录日志中留下零个非 GET 请求;同时(正向对照)apply_proposal 确实会在同一 fixture 中记录该文档所述的写入调用。--live-probe 是一个诊断标志(发出单个真实请求,用于区分“端点正确但未授权”和“端点错误”);它绝不会在测试或 CI 中运行,且需要设置 LINKEDIN_MCP_LIVE_PROBE=1。
API 面固定为官方文档中带日期的逐字摘录,参见 docs/api-notes.md。
许可证
MIT——参见 LICENSE。
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- AlicenseAqualityDmaintenanceEnables Claude to interact with LinkedIn for posting, commenting, sharing media, and viewing feed and profile.91773MIT
- AlicenseNot gradedqualityDmaintenanceEnables Claude AI to interact with LinkedIn through browser automation, including profile reading, people and job search, company research, post publishing, and profile editing.MIT
- AlicenseAqualityDmaintenanceEnables posting content, sharing articles, and managing your LinkedIn profile through Claude using the official LinkedIn API with OAuth 2.0.610MIT
- AlicenseBqualityCmaintenanceEnables drafting, formatting, analyzing, and publishing LinkedIn posts directly from MCP-compatible clients like Claude.5177MIT
Related MCP Connectors
233 tools for Google, Microsoft, TikTok, LinkedIn Ads in Claude or ChatGPT. Writes need approval.
Let AI tools securely access your LinkedIn network and DMs
Live SEO workflow tools for Claude Code, Codex, and AI agents.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/kyle-nelson-berkeley/linkedin-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server