AI Interview Agents——MCP Server is the official Model Context Protocol server for AI Interview Agents (a voice AI recruiting platform). Its purpose is to let any MCP client (such as Claude, Cursor, or a custom agent) drive the entire recruiting workflow using natural language.
Main capabilities (according to the material, the server exposes the following functions):
- Roles: List your roles, or create a new role based on a job description.
- Candidates & CVs: Parse uploaded CV/PDF files; normalize messy pasted candidate lists (name, email, phone, experience, skills); reassign candidates between roles.
- Interviews: Schedule interviews (providing a dry_run mode, which only normalizes and previews, sends nothing, and consumes no quota), reschedule, cancel, and send reminders.
- Reports: View candidate pipeline status, and pull scored interview reports with transcripts.
Scope note to be aware of: the actual live voice interview is not within the scope of this server; it runs on a separate real-time service. This server is responsible for triggering interviews and reading back results.
How to use:
- Hosted server (recommended): The server runs remotely and uses OAuth login. Point any MCP client that supports Streamable-HTTP at the service, and the client will automatically discover the authentication flow via /.well-known/oauth-protected-resource and log in with your AI Interview Agents account. The server is also registered in the official MCP Registry under the identifier com.aiinterviewagents/interviews.
- Local run (stdio mode): Local deployment is supported, and AIIA_BACKEND_URL can be used to override the API base address to point to your own deployment.
Security and trust model: The server itself does not store account secrets. Each request carries the user's own Firebase ID token and forwards it to the backend as a bearer token, using the same trust path as the web app. For hosted transport, through the OAuth layer (dynamic client registration + PKCE) connects the MCP client with Firebase login.
Use cases: Users on recruiting teams or in recruiting workflows who want to complete steps such as role creation, resume screening, interview scheduling, and result retrieval using natural language through an MCP client.
Limitations and information not provided:
- Live voice interviews are performed by a separate real-time service; this server does not include that functionality.
- The material does not provide specific pricing, costs, supported platform details, quota rules, performance, or specific deployment step examples.
- The license is MIT; see the LICENSE file in the repository for details.
- The material does not provide language support information other than English, Hindi, and other Indian languages.