Developer access
Remote jobs from several boards, in one place, reachable by something other than a browser. No key, no account.
MCP server
Point any MCP client — Claude, an editor, your own agent — at the endpoint below. It speaks Streamable HTTP and is stateless, so there is no session to manage.
{
"mcpServers": {
"anywhereroles": {
"type": "http",
"url": "https://anywhereroles.com/mcp"
}
}
}Eight tools:
search_jobs— Search listings by keyword, category, source and salary rangeget_job— Full detail for one listing by idlist_categories— Every job categorylist_tags— Every skill and technology tagget_company— One company: job count, sources, top categories and tagslist_companies— Search the company directoryget_statistics— Aggregate market statistics across every sourceget_salary_insights— Salary ranges by category and by source
Two resources are also exposed: remoti://jobs/latest and remoti://statistics.
Job feeds
Any search on this site is also an RSS feed. Take the filters off a search URL and put them on /feed.xml — they are the same parameters.
- /feed.xml — Everything, newest first
- /feed.xml?q=react — Anything matching a keyword
- /feed.xml?categoryId=3 — One category
- /feed.xml?sources=Himalayas — One source board
- /feed.xml?q=react&salaryMin=120000 — A keyword above a salary floor
Feeds are always newest-first: a feed pinned to “highest salary” would never change, and so would never deliver anything.
Attribution
Every listing carries the name of the board it came from and a link to the original posting. Keep both when you display them. Several sources require the credit and the link back, and several prohibit their listings being submitted to other job platforms — so please read and link, rather than re-list.
Questions
Do I need an API key or an account?
No. Both the MCP endpoint and the feeds are public and unauthenticated. Requests are rate-limited per IP, so an agent loop should pace itself rather than retry immediately.
What data comes back?
Job metadata — title, company, location, employment type, salary, category, tags — along with the source board's name and the canonical link to the original listing. Full job descriptions are not served: they belong to the boards we aggregate, and several of their terms only permit metadata plus a link back.
Can I republish these listings on another job board?
No. Several of the sources we aggregate prohibit submitting their listings to third-party job platforms by name. These surfaces are for reading and linking, not for re-listing.
How fresh is it?
Feeds are cached for an hour. The MCP endpoint reads live. Every job carries its publication date, and each source page states when that board last published anything.