Techmap Job Postings

What Techmap Job Postings does

The Techmap Job Postings MCP server exposes Techmap’s international job-posting data to MCP-compatible AI assistants and agents. The data covers approximately 8 million new postings per month, sourced from company career pages, ATS platforms, job boards, aggregators, and public employment offices. Historical data is available from 2020, according to the project documentation.

Search results can be narrowed by country, job title, occupation, skills, company, city, workplace arrangement, contract or work type, language, industry, posting date, salary, and whether the listing comes directly from an employer. Search responses contain 10 postings per page.

The server belongs to the 🔎 Search & Data Extraction category and is suitable when an agent needs current or historical vacancy data rather than a general web search.

How it works

The server runs through the @techmap/mcp-server npm package and connects to the Techmap Jobs API through RapidAPI. An MCP client starts the server with npx, while the RapidAPI credential is supplied as an environment variable. API usage is therefore tied to the RapidAPI subscription associated with that key.

Four MCP tools are documented:

• search_jobs retrieves postings matching the selected filters. Results are paginated at 10 postings per page.
• count_jobs returns the number of postings matching the same filter set, which supports labour market questions and volume comparisons.
• list_filter_values reports frequent values for fields such as workplace, industry, occupation, skills, company, and city over the previous 30 days. It can help identify exact values for later searches, as well as common companies or skills. This tool requires a PRO plan or higher.
• get_rss_feed_url creates a filtered RSS feed URL for job-board backfill. The generated URL includes a key placeholder.

Setup and configuration

The Techmap Job Postings MCP server requires a RapidAPI key and an active subscription to the Techmap Jobs API. The documented BASIC plan includes 1,000 job postings per month with a limit of one request per second. Paid usage is listed as $1 per 1,000 postings.

For Claude Desktop, Claude Code, or Cursor, add an MCP server entry that runs npx -y @techmap/mcp-server and sets TECHMAP_RAPIDAPI_KEY to the RapidAPI key. Claude Code can also add the server from the command line with claude mcp add. Smithery provides an installation path that asks for the RapidAPI key during setup.

The key remains on the local machine and is sent to RapidAPI by the server. The repository also documents local development commands using npm install, npm test, and the MCP Inspector.

Tools and capabilities

The Techmap Job Postings MCP server supports both record retrieval and aggregate analysis. Use search_jobs when the agent needs individual vacancies, such as remote Python roles in a particular country. Use count_jobs when the question asks how many postings match a time, location, occupation, or other combination of filters.

list_filter_values is useful before searching when the required spelling of a company, occupation, skill, or city is uncertain. Because it examines the most common values from the last 30 days, it is oriented toward recent filter discovery rather than unrestricted historical lookup.

For workflows that maintain an external job board, get_rss_feed_url can produce a filtered feed URL instead of returning postings directly. The documentation describes this as a backfill-oriented capability.

Limitations and notes

The server depends on RapidAPI credentials, subscription limits, and request restrictions. The BASIC plan has a monthly posting allowance and a one-request-per-second limit. search_jobs returns 10 postings per page, so larger result sets require additional page requests.

list_filter_values is not available on every plan; the README specifies a PRO plan or higher. The RSS tool builds a URL with a key placeholder, so downstream feed use requires the appropriate key handling.

The Techmap Job Postings MCP server is focused on the data and tools documented by the repository. The project also links to separate API documentation, country-level data feeds, historical datasets, and sample datasets, but those resources are not additional MCP tools exposed by this server.