Greenhouse MCP

Greenhouse's MCP server connects AI assistants to an applicant tracking system, so recruiting work can happen in a conversation instead of across dozens of pipeline views. An agent can look up jobs and their open requisitions, search candidates and applications, read the stage a candidate sits in, pull interview schedules and scorecard feedback, and retrieve pipeline reporting such as stage conversion and time in process. Permissions are the defining constraint. The server authenticates as a specific Greenhouse user or scoped credential and honors that user's role and job level access, so a coordinator and a hiring manager see different slices of the same workspace. That matters because applicant tracking data includes personal information and regulated hiring records, and an integration that ignored role boundaries would be unusable for most talent teams. It is built for recruiters, sourcers, hiring managers and talent operations teams who want an assistant to answer questions like which candidates are stuck in a stage, what feedback an interview loop produced, or how a requisition is trending, and for people ops engineers building internal tooling on top of Greenhouse. As a standard MCP server it plugs into Claude and other compatible clients, letting recruiting context sit alongside whatever else the assistant is working on.