@gera-services/mcp-gera-jobs

GeraJobs MCP server — give any AI agent (Claude, ChatGPT with tools, Cursor, or any Model Context Protocol client) direct access to the GeraJobs global jobs dataset: 30 job categories, 20 country job markets, and salary benchmarks aggregated from 1,002 real crawled job listings.

Everything is answered locally from on-disk data — no backend, no network, no auth, no API key. It is the same data the GeraJobs product renders. A Gera Systems product.

Tools

| Tool | What it does | |---|---| | search_job_categories | Free-text search over the 30 categories (name, description, skills, keywords). Returns avg USD salary range, top skills, remote %, trending flag. | | get_salary_benchmark | Real salary benchmarks for a role per country/currency — sample size + median / p25 / p75 / low / high annual pay, plus sample listings. Aggregated from crawled listings. | | list_job_categories | Compact index of all 30 categories (id, name, salary range, trending) — the canonical ids for the other tools. | | get_job_market | Country hiring profile: avg monthly wage, top industries/employers, job + AI growth, remote adoption, summary. Lookup by ISO code, country name, or city. | | get_category_details | Full profile + curated FAQ for one category (salaries, skills, how to find roles). |

All numbers trace to the underlying GeraJobs datasets; salary figures derive from real crawled listings (snapshot date and total listing count are returned in every salary response). Orientation only — not a contractual salary guarantee.

Install & run

# From the npm registry (once published):
npx -y @gera-services/mcp-gera-jobs

# Or from this repo:
cd packages/mcp-gera-jobs
node scripts/build-dataset.mjs   # snapshot the real datasets (already committed)
npm run build                    # tsc --noCheck -> dist/
npm run smoke                    # prove tools/list + tools/call over stdio

Claude Desktop

Add this to your claude_desktop_config.json (~/Library/Application Support/Claude/claude_desktop_config.json on macOS):

{
  "mcpServers": {
    "gera-jobs": {
      "command": "npx",
      "args": ["-y", "@gera-services/mcp-gera-jobs"]
    }
  }
}

Then restart Claude Desktop. Ask things like:

• "What's the median salary for a data scientist in the US?"
• "Show me trending remote job categories and their top skills."
• "What's the hiring market like in Armenia right now?"

Cursor / other stdio MCP clients

Point the client at the mcp-gera-jobs stdio command (npx -y @gera-services/mcp-gera-jobs), or to node bin/cli.js when running from a local checkout.

Data sources (real, in-repo)

• Categories: apps/gerajobs-web/src/data/job-categories.ts (30 categories)
• Markets: apps/gerajobs-web/src/data/job-markets.ts (20 countries)
• Salary benchmarks: apps/gerajobs-web/src/data/salary-cluster.generated.ts — aggregated from 1,002 real crawled GeraJobs listings.

Regenerate the bundled snapshot with node scripts/build-dataset.mjs.

A GeraJobs product, by Gera Systems Ltd. MIT licensed.