Careerjet MCP Server
Pricing
Pay per event + usage
Careerjet MCP Server
Query Careerjet jobs live from Claude, ChatGPT, Cursor and other AI agents via MCP. Search jobs, get counts, resolve locations and autocomplete keywords across 90+ countries — or run it as a classic scraper. Pay only per tool call.
Give your AI assistant live access to the global job market. This is a Model Context Protocol (MCP) server for Careerjet — a job-search aggregator indexing millions of postings across 90+ countries — so agents like Claude, ChatGPT, Cursor, and n8n can search jobs, size demand, resolve locations, and suggest keywords in real time.
It also runs as a classic scraper: start a normal run with the input schema and it pushes normalized job rows to a dataset.
Why AI teams use it
- Recruiting copilots — "Find remote data-engineer roles in Germany posted this week" answered live in chat.
- Market-intel agents — compare demand for a role across cities or countries
with
count_jobs. - Job-search assistants — resolve a vague location, autocomplete keywords, then return matching openings.
MCP tools
| Tool | What it does |
|---|---|
search_jobs | Search live jobs by keywords, location, country, and filters |
count_jobs | Return the total number of matching jobs (market sizing) |
resolve_location | Turn free-text into Careerjet's location tree with job counts |
autocomplete_keywords | Suggest search keywords for a partial query |
Each job row includes title, company, salary, location, contract type, posting date, and the full description.
Connect it
- Start the actor in Standby mode (it exposes an MCP endpoint).
- Point your MCP client at
<ACTOR_STANDBY_URL>/mcp(Streamable HTTP). - In Claude Desktop, Cursor, or your agent framework, add it as an MCP server — the four tools appear automatically.
Example tool call
{"tool": "search_jobs","arguments": {"keywords": "registered nurse","location": "Texas","country_code": "US","sort": "date","limit": 20}}
Classic scraper mode
Prefer a dataset? Run it normally with:
{"keywords": "software developer","location": "New York","countryCode": "US","maxItems": 100}
Rows land in the dataset with the same fields, ready for JSON/CSV/Excel export or Google Sheets, Make, and Zapier.
Output fields
job_id, title, company, company_logo_url, location, locations, salary, contract_type, contract_period, is_new, posted_at, snippet, description_html, description_text, url, country_codeFAQ
Which countries? 90+ — pass country_code (US, GB, DE, FR, IN, AU, …).
Do I need credentials? No — Careerjet's job API is open. No proxy needed.
How fresh is the data? Pulled live at call time; use sort: "date" for the
newest postings.
Is it reliable? It calls Careerjet's own JSON API instead of scraping HTML, so it doesn't break when the website changes.
Can I still use it without MCP? Yes — run it as a normal scraper via the input schema.