The Talroo Search MCP is a specialized integration for AI agents and chatbots. Most publishers should use the API, XML, widget, or affiliate link integrations instead.
Talroo's Job Search MCP (for AI Agents)
Talroo has an HTTP MCP server so AI tools and agents can search jobs with the same id and pass as the Search API.
MCP (Model Context Protocol) is a standard way to connect tools to external data. You don't have to build a custom integration for each client.
Setup
| Property | Value |
|---|---|
| Server type | HTTP, using Model Context Protocol |
| Authentication | Same as Search API (username and password) |
| URL | https://mcp.talroo.com/mcp?id=YOUR_API_ID&pass=YOUR_API_PASS |
Configuration for AI Tools
For Cursor and similar tools, add this to the MCP config file:
{
"talroo": {
"type": "http",
"url": "https://mcp.talroo.com/mcp?id=YOUR_API_ID&pass=YOUR_API_PASS"
}
}
Using the MCP
The main tool is job_search. Parameters:
| Parameter | Required | Description |
|---|---|---|
location | Yes | <City,state>, or zipcode |
query | No | User query |
ip | No | Location resolution and security/privacy considerations |
Optional parameters can change. Check this page again if you pin them.
Health Check
If you are having trouble accessing the server, check the server health via:
https://mcp.talroo.com/health
Example Request and Response
Wire it up like any other MCP server. Details depend on the agent.
Request:
{
"method": "tools/call",
"params": {
"name": "job_search",
"arguments": {
"query": "Python developer",
"location": "Austin, TX"
}
}
}
Response:
{
"jobs": [
{
"title": "Python Developer",
"company": "TechCorp",
"location": "Austin, TX",
...
}
],
"total_results": 156,
"page": 1,
"search_metadata": { ... }
}
Call it asynchronously, retry with backoff, and don't assume every tool call succeeds.
In chat, agents aren't guaranteed to pick the right tool. A prompt such as "Search for Python jobs in Austin using the Talroo MCP" helps.