jobs-mcp-server

MCP server that exposes your MongoDB healthcare jobs collection to any MCP-compatible client — ChatGPT (via the OpenAI MCP connector), Claude Desktop, or any other MCP host.

---

Tools available

Tool	What it does
list_jobs	Paginated list with filters (status, job_type, client_name, is_active)
get_job	Fetch a single job by _id
search_jobs	Full-text search across title, client, location, description, notes
create_job	Insert a new job
update_job	Patch any fields on an existing job
delete_job	Soft-delete (is_active=false) or hard-delete
jobs_summary	Dashboard counts by status and job type

---

Setup

1. Install dependencies

cd jobs-mcp-server
npm install

2. Configure environment

Create a .env file or export the variables before running:

export MONGO_URI="mongodb://localhost:27017"   # your MongoDB connection string
export DB_NAME="xpress_health"                 # your database name
export COLLECTION="jobs"                       # collection name (default: jobs)

For MongoDB Atlas:

export MONGO_URI="mongodb+srv://user:password@cluster.mongodb.net/?retryWrites=true&w=majority"

3. Run the server

node server.js

The server communicates over stdio (standard MCP transport).

---

Connect to ChatGPT

ChatGPT supports MCP servers via the OpenAI MCP remote connector.
Since ChatGPT requires an HTTP/SSE endpoint (not raw stdio), you need to wrap the server with a small HTTP bridge.

Step 1 – Install the SSE bridge

npm install -g @modelcontextprotocol/server-stdio-to-sse
# or use npx without installing globally

Step 2 – Run the bridge

npx @modelcontextprotocol/server-stdio-to-sse \
  --port 3100 \
  -- node /absolute/path/to/jobs-mcp-server/server.js

This starts an SSE HTTP server on http://localhost:3100.

Step 3 – Expose it publicly (for ChatGPT to reach it)

ChatGPT needs a public HTTPS URL. Use ngrok or any tunnel:

# Install ngrok: https://ngrok.com/download
ngrok http 3100
# → gives you: https://abc123.ngrok-free.app

Step 4 – Add to ChatGPT

1. Go to ChatGPT → Settings → Beta Features → MCP Servers (or the Connectors panel)
2. Click Add Server
3. Enter the URL: https://abc123.ngrok-free.app/sse
4. Give it a name: Healthcare Jobs
5. Save — ChatGPT will auto-discover the 7 tools

Now you can ask ChatGPT:

• "List all pending jobs"
• "Search for wound care jobs in Cork"
• "Create a new personal care job for Mary Delaney"
• "Give me a summary of all jobs by status"

---

Connect to Claude Desktop

Add this to your claude_desktop_config.json:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "jobs": {
      "command": "node",
      "args": ["/absolute/path/to/jobs-mcp-server/server.js"],
      "env": {
        "MONGO_URI": "mongodb://localhost:27017",
        "DB_NAME":   "xpress_health",
        "COLLECTION": "jobs"
      }
    }
  }
}

Restart Claude Desktop — the jobs tools will appear automatically.

---

Connect to Cursor / Windsurf / VS Code

Add to your MCP config (.cursor/mcp.json or equivalent):

{
  "mcpServers": {
    "jobs": {
      "command": "node",
      "args": ["/absolute/path/to/jobs-mcp-server/server.js"],
      "env": {
        "MONGO_URI": "mongodb://localhost:27017",
        "DB_NAME": "xpress_health"
      }
    }
  }
}

---

Production deployment (HTTPS without ngrok)

Deploy the SSE bridge on any Node.js host (Railway, Render, Fly.io, VPS):

# On your server
MONGO_URI="mongodb+srv://..." DB_NAME="xpress_health" \
npx @modelcontextprotocol/server-stdio-to-sse \
  --port 3100 \
  -- node /app/jobs-mcp-server/server.js

Point your domain / reverse proxy at port 3100 with HTTPS, then use that URL in ChatGPT.

---

Example ChatGPT prompts once connected

List all jobs scheduled this week
Show me all In Progress jobs for wound care
Create a job: Morning Personal Care for Patrick Cronin, Personal Care type, scheduled 2026-05-20
Mark job 683abc... as Completed
How many jobs are in each status?
Search for jobs in Kerry
Delete job 683abc... (soft delete)