# API and MCP connectivity

Intervieux exposes a full ATS REST API, covering jobs, applications, pipeline, scheduling, documents, and talent search across 13 scopes, alongside a hosted remote MCP server with roughly 40 tools and a real OAuth consent screen, so Claude, ChatGPT, or your own internal tools can connect directly to your live hiring data rather than working from an export.

Two different problems get solved by the same underlying access here. One is pulling your ATS into a tool you already use, a dashboard, a spreadsheet script, an internal app, through a documented REST API with an OpenAPI spec. The other is letting an AI tool you already trust, like Claude or ChatGPT, act on your hiring data directly through MCP, with a consent screen that shows exactly what it's asking permission to do before you grant it. Neither path requires waiting on a bespoke integration built specifically for your setup; both are the same standard surface every integration works against.

## What connectivity covers

The ATS v1 REST API covers jobs, applications, pipeline, scheduling, and documents, plus talent search, split across 13 scopes so an integration only gets access to what it actually needs, with the full surface documented as an OpenAPI spec at /api/v1/. The hosted remote MCP server at app.intervieux.ai/mcp runs on OAuth 2.1 with PKCE and Dynamic Client Registration, meaning a tool like Claude or ChatGPT can request a connection and you approve it through a real consent screen, not a shared API key pasted into a config file. Around 40 MCP tools are exposed through that connection. Generic inbound MCP connectors extend the same pattern to the employer agent itself, and separately, google_workspace_cli and microsoft_365_cli let the agent act through your own connected Google Workspace or Microsoft 365 account, for things like checking a real calendar before proposing an interview time.

## A walkthrough: connecting Claude to the pipeline directly

An operations lead wants to ask Claude Desktop questions about hiring without switching tabs. They add Intervieux's MCP server URL as a connector, and Claude opens a real OAuth consent screen listing the specific scopes being requested, applications and pipeline read access, nothing broader. They approve it. From then on, asking Claude "how many candidates are in screening for the backend role" runs through the MCP connection using the same underlying data the ATS itself reads from, no export, no copy-paste. Separately, inside the Intervieux agent itself, the employer connects their Google Workspace account, and when they later ask the agent to check for real calendar conflicts before proposing interview times, it uses google_workspace_cli to check the actual calendar rather than assuming an availability window is open.

## What connectivity does and doesn't grant

Every MCP connection goes through a real consent screen naming the scopes being requested, using OAuth 2.1 with PKCE, not a static key that grants blanket access, and Dynamic Client Registration means a new connecting tool doesn't need a manually issued credential handed to it first. The REST API is split into 13 scopes specifically so an integration can be limited to what it actually touches, jobs, applications, or talent search, rather than an all-or-nothing credential that could reach everything the moment it's issued. Connected Google Workspace and Microsoft 365 accounts are used by the agent to check or act through tools you've explicitly connected, like a calendar, rather than assumed by default. No native Greenhouse, Lever, or Workday connector exists; the REST API and MCP server are the integration surface, not a pre-built sync to another ATS.

## Frequently asked questions

### Can I connect Claude or ChatGPT directly to my Intervieux hiring data?

Yes, through the hosted remote MCP server at app.intervieux.ai/mcp, which uses OAuth 2.1 with PKCE and a real consent screen, exposing roughly 40 MCP tools once a connection is approved.

### Is there a documented REST API for pulling ATS data into my own tools?

Yes, a full ATS v1 REST API covering jobs, applications, pipeline, scheduling, documents, and talent search, split across 13 scopes, with an OpenAPI spec published at /api/v1/.

### Does connecting an AI tool through MCP give it unrestricted access to my data?

No, the consent screen names the specific scopes being requested before you approve the connection, and the REST API's 13-scope structure limits what any single integration can reach.

### Can the agent check my real Google or Microsoft calendar before proposing interview times?

Yes, through connected Google Workspace or Microsoft 365 accounts and their respective CLI tools, the agent can check real calendar data rather than assuming a window is open.

### Does Intervieux have a native integration with Greenhouse or Lever?

No, no native connector to those systems exists. The REST API and MCP server are the way to connect Intervieux to another system or tool.

## Related pages

- [Your AI recruiting agent](/recruiting-agent)
- [Scheduling automation](/recruiting-agent/scheduling-automation)
- [Hiring funnel analysis](/recruiting-agent/hiring-analytics)
- [Scheduled recruiting tasks](/recruiting-agent/scheduled-tasks)
- [Frequently asked questions](/faq)

## Connect your own tools to the pipeline

Point an MCP-capable AI tool at the hosted connector, or build against the documented REST API for your own integration.

Start practicing free: https://www.intervieux.ai/register · Hire with Intervieux: https://www.intervieux.ai/employers/signup