Jobo Job Search MCP server

Jobo Job Search MCP Server

Remote MCP server exposing Jobo's live job index — millions of listings collected from employer career sites and 150+ applicant tracking systems — to LLM clients.

Full client setup (Claude, ChatGPT, Cursor, Codex CLI) and the tool reference: jobo.world/docs/connectors/mcp.

natively and every 2025-era client through the built-in legacy fallback (SDK v2 createMcpHandler, one factory for both eras).

• Transport: Streamable HTTP, single /mcp endpoint, stateless. Serves MCP spec 2026-07-28

with their Jobo account — no API key copy-paste. Required scope: jobs:read.

• Auth: OAuth 2.1. This is a Resource Server; the Authorization Server is the Jobo API. Clients log in

Self-hosting

The hosted deployment is https://jobs-mcp.jobo.world. To run your own copy instead:

npx jobo-job-search-mcp

Starts the Streamable HTTP server on $PORT (default 3002); point your client at http://localhost:3002/mcp. This changes where the gateway runs, not its auth model — it's still an OAuth resource server gated on Jobo account sign-in, since the upstream API validates every request against the Authorization Server regardless of which copy of the gateway forwarded it. Set MCPRESOURCEURL to match whatever host you actually serve it from — see Configuration below.

Why this is a separate server

Jobo.Enterprise/Jobo.Enterprise.Mcp was deliberately re-scoped to analytics-only in v4, which removed searchjobs, getjobdetails, listfilters, search and fetch. Adding job tools back there would undo that decision, so this is a second server against the same External API.

The /api/mcp/jobs/* endpoints were never removed — McpController.cs still serves them, and its own comment notes the GET search is "convenient for the canonical ChatGPT search(query) tool". This server is a thin OAuth-forwarding gateway in front of endpoints that were built for it.

The immediate payoff: search + fetch restore Deep Research compatibility. Without that canonical pair a server cannot be used as a ChatGPT Deep Research connector at all.

Tools

search/fetch deliberately take the minimum arguments the contract allows. Anything with structure should go through search_jobs, where filters are real parameters rather than hopeful free text.

What fetch returns

text is self-contained prose, because Deep Research reads it and never opens the URL. It is built from the AI-extracted fields (responsibilities, qualifications, benefits, compensation) in preference to the raw employer HTML, which is boilerplate-heavy and frequently longer than it is useful. The raw description is available via getjobdetails with include_description: true.

Auth model

The server is a gateway, not the cryptographic authority. The C# External API validates the JWT with OpenIddict against the same issuer, audience and jobs:read scope; verifying the signature a second time here would only let the two validators drift. So this does the minimum a gateway must:

Shortened. The full README is on GitHub.