Remote Jobs Center
Search current remote / work-from-home jobs by category, region, perk, or company — with ready-to-apply links.
README
Remote Jobs Center MCP Server
A hosted Model Context Protocol server for searching current remote / work-from-home jobs — by category, region, perk, or company. Built so any MCP-enabled assistant or agent can query the Remote Jobs Center corpus directly and return ready-to-apply links.
Unlike single-board servers, the corpus is aggregated and deduped from multiple vetted remote job boards and employer career sites, then enriched with perks, skills, seniority, and AI-role flags — so a single query spans the whole remote market, not one source.
- Endpoint:
https://remotejobscenter.com/mcp - Transport: Streamable HTTP (JSON-RPC 2.0 over POST) — stateless, no auth
- Registry:
com.remotejobscenter/remote-jobs(official MCP registry)
Connect
Claude (web/desktop): Settings → Connectors → Add custom connector → URL https://remotejobscenter.com/mcp. No authentication required.
Any MCP client that supports Streamable HTTP: point it at the endpoint above.
Tools
| Tool | Parameters | Returns |
|---|---|---|
search_remote_jobs |
query, category, region, perk, company, limit (all optional) |
{ total, returned, jobs[] } |
get_job |
slug (required) |
{ job } or { error } |
list_categories |
— | { categories: [{ category, count, url }] } |
list_collections |
— | { collections: [{ perk, count, url }] } |
list_companies |
category (optional) |
{ companies: [{ company, count }] } |
Filter values
category:software,data,design,marketing,sales,support,writing,product,finance,hrregion:us,uk,europe,canada,latam,apac,worldwideperk:tuition-reimbursement,six-figure,no-degree,entry-level,four-day-week,visa-sponsorship,async,equity,unlimited-ptolimit: default20, max50
Job shape
{
"slug": "acme-senior-backend-engineer",
"title": "Senior Backend Engineer",
"company": "Acme",
"category": "software",
"region": "us",
"location": "United States (Remote)",
"salaryMin": 140000,
"salaryMax": 180000,
"perks": ["six-figure", "equity"],
"source": "Employer careers",
"postedAt": "2026-06-10T00:00:00.000Z",
"applyUrl": "https://...",
"url": "https://remotejobscenter.com/jobs/acme-senior-backend-engineer/"
}
get_job additionally includes a short excerpt.
Request / response examples
All requests are JSON-RPC 2.0 over POST https://remotejobscenter.com/mcp.
Initialize
curl -s https://remotejobscenter.com/mcp \
-H 'Content-Type: application/json' \
-d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2024-11-05"}}'
List tools
curl -s https://remotejobscenter.com/mcp \
-H 'Content-Type: application/json' \
-d '{"jsonrpc":"2.0","id":2,"method":"tools/list"}'
Search
curl -s https://remotejobscenter.com/mcp \
-H 'Content-Type: application/json' \
-d '{"jsonrpc":"2.0","id":3,"method":"tools/call","params":{"name":"search_remote_jobs","arguments":{"query":"python","region":"us","limit":5}}}'
A tools/call result wraps the tool's JSON output in a text content block:
{ "jsonrpc": "2.0", "id": 3, "result": {
"content": [ { "type": "text", "text": "{\"total\":42,\"returned\":5,\"jobs\":[...]}" } ]
} }
Run locally (stdio)
The hosted endpoint above is the canonical deployment. For local use or registry introspection, this repo also ships a standalone stdio MCP server with the same tools and data source.
npm install
npm start # runs src/server.mjs over stdio
Point any stdio MCP client at node /path/to/remotejobscenter-mcp/src/server.mjs.
Docker
docker build -t remotejobscenter-mcp .
docker run -i --rm remotejobscenter-mcp
The container speaks MCP over stdio, so registries (e.g. Glama) and MCP clients can start it and introspect its tools directly.
Implementation
Two equivalent surfaces, one data source:
- Hosted — a single Cloudflare Pages Function deployed at
/mcp, speaking Streamable HTTP (JSON-RPC 2.0 over POST). Stateless, read-only, no auth. AGETwithAccept: text/event-streamreturns405(per the Streamable HTTP spec, since the server does not stream); a plainGETreturns a small info document. - Local / containerized —
src/server.mjs, a stdio MCP server using the official@modelcontextprotocol/sdk.
Both read the same public corpus the website publishes at
https://remotejobscenter.com/api/jobs.json (cached ~10 minutes), so there is one source of truth.
License
MIT — see LICENSE.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。