MCP server

Add it to your assistant

claude mcp add --transport http 4dayweek https://4dayweek.io/api/mcp

https://4dayweek.io/api/mcp

https://4dayweek.io/api/mcp

{
  "mcpServers": {
    "4dayweek": {
      "url": "https://4dayweek.io/api/mcp"
    }
  }
}

{
  "servers": {
    "4dayweek": {
      "type": "http",
      "url": "https://4dayweek.io/api/mcp"
    }
  }
}

curl -X POST https://4dayweek.io/api/mcp \
  -H 'Content-Type: application/json' \
  -H 'Accept: application/json, text/event-stream' \
  -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{
        "protocolVersion":"2025-06-18","capabilities":{},
        "clientInfo":{"name":"curl","version":"1.0"}}}'

Tools

6 tools. Every one is read-only and says so in the protocol, so a cautious client can call them freely.

Try asking

• “Find me remote 4-day-week engineering jobs open to applicants in the UK.”
• “What is a 9-day fortnight, and is it the same as a compressed week?”
• “Which employers run a genuine 4-day week at full pay, and how many are there?”
• “Tell me about the 4-day week at a specific company, and how it was verified.”
• “Show me senior product roles at 32 hours a week, posted in the last fortnight.”

Limits and guarantees

• Read-only, by construction. The server reaches the database through interfaces that expose only read methods, so a tool that writes would not compile. Nothing here can change anything.
• No personal data. The tools return public job listings and employer profiles. There is no account to connect and no user data to reach.
• Paging. Search tools return at most 25 results per call and take a page argument up to 200. Each response carries has_more and, where there is one, next_page.
• Rate limit. 300requests per minute per IP, separate from the REST API’s own bucket. Over the limit returns HTTP 429 with a Retry-After header.

Data and attribution

Listings come from employers’ own career sites and applicant tracking systems, and each employer’s schedule is recorded against the criteria on our methodology page. Aggregate market data is published under CC BY 4.0 on the market data page. If you build something on this, please credit 4dayweek.io and link back.

Prefer plain HTTP? The same listings are available through the public REST API.

Questions

Do I need an API key?

No. The server is read-only and anonymous. There is no account, no key and no OAuth step — point your client at the URL and it works.

What can it do to my data?

Nothing. Every tool is read-only and declares it in the protocol, so a cautious client can call them freely. The server has no write path at all: the code can only reach read methods, which is enforced by the compiler rather than by review.

Will it tell an assistant that a 4x10 is a 4-day week?

No, and that is the main reason to use it rather than scraping. A compressed week is the same 40 hours in four days. Every listing carries an explicit schedule type, the tool descriptions define each one, and explain_schedule_type states whether a schedule genuinely reduces hours and whether that reduction is at full pay.

How many results can I get?

Search tools return at most 25 results per call and accept a page argument up to 200. Each response says whether another page exists.

What are the rate limits?

300 requests per minute per IP, separate from the REST API's own limit. Going over returns HTTP 429 with a Retry-After header.

Can I use the data in my own product?

Yes. Please credit 4dayweek.io and link back. Aggregate market data is published under CC BY 4.0 on our data page.

How does someone actually apply to a job?

Every result carries the job's page on 4dayweek.io, and applying happens there. The tools never hand out an employer's raw application URL, so the employer keeps seeing where the applicant came from.