Metaview MCP Server

Metaview is an AI recruiting and interview intelligence platform that automatically captures and structures interview conversations, candidate notes, and hiring funnel data. The official Metaview MCP server exposes this captured interview intelligence to MCP-compatible AI clients (Claude is called out as the primary supported client) so recruiters and hiring managers can analyze candidates, compare pipelines, and run outreach using natural language instead of building reports or exporting data.

The integration spans three core Metaview product areas: Notes (querying interview transcripts and scorecard feedback), Reports (hiring funnel analytics, time-to-hire, compensation), and Sequences (creating and managing candidate outreach). Once connected, Claude can answer questions like "compare the top three candidates for this role" or "which roles have high interview volume but low offer rates" against your live Metaview workspace.

Setup is handled inside Metaview itself rather than via a self-hosted package: admins enable the MCP integration in workspace settings, then any authorized user can connect from Settings → MCP and link Claude in roughly two minutes. There is no public GitHub repository, no npm package, and no custom config JSON to write; the server is hosted by Metaview and uses the in-app connector flow.

Tools

Prerequisites

• An active Metaview workspace with interview data
• A workspace admin to enable the MCP integration (non-admins can be granted access via a separate workspace setting)
• An MCP-compatible client (Claude is the officially supported client; other MCP clients that support remote servers should also work)

Enable MCP on your workspace

1. Sign in to Metaview as an admin.
2. Click the gear icon in the top left to open Workspace Settings.
3. Under Integrations, enable MCP Integration. Optionally enable MCP for Non-Admins if you want all members to be able to connect.

Connect from Metaview

1. In Metaview, go to Settings → MCP.
2. Follow the in-app instructions to authorize an AI client (Claude) against your Metaview workspace.
3. The connection completes the OAuth-style handshake. Setup is documented as taking about two minutes and does not require manual config files.

Notes

• Metaview describes this as the "official Claude connector" with no custom config needed, so the server URL and credentials are issued through the in-app flow rather than copied into a claude_desktop_config.json.
• There is no public GitHub repo, npm package, or self-hosted binary for this server. It is provider-hosted by Metaview.
• For a guided setup, Metaview offers a live demo at https://www.metaview.ai/demo/mcp.

Alternative: Zapier MCP

If you do not have direct MCP enabled on your plan, Metaview is also exposed through Zapier's MCP server, which provides a smaller set of actions (find conversation, find conversation property, create or update conversation property). See https://zapier.com/mcp/metaview.

• Compare the top candidates for an open role using interviewer notes and scorecard feedback without manually opening each interview report.
• Investigate hiring funnel drop-off: ask why time-to-hire is increasing or which stage has the highest rejection rate for a given role.
• Surface recurring red flags or skill gaps that interviewers raise across an entire pipeline.
• Run compensation analysis across roles and geographies (e.g. average expected salary by role and region, remote vs hybrid).
• Spin up or update candidate outreach sequences from a conversation with Claude, including adding candidates and editing email content.

• "Compare the top three candidates for the Senior Backend Engineer role and summarize each candidate's strengths and concerns."
• "What are the common concerns interviewers raised for the Product Manager pipeline this quarter?"
• "Which open roles have high interview volume but low offer rates in the last 60 days?"
• "What is the average salary expectation for Staff Engineer candidates across the US and UK?"
• "Create an outreach sequence for senior data scientists and add the five candidates I just interviewed last week."

• Official, provider-hosted server built and maintained by Metaview, so it stays in sync with the underlying product.
• Covers three meaningful surfaces (Notes, Reports, Sequences) rather than just read-only access to a single data type.
• Quick setup with no config file editing; auth is handled through the Metaview workspace.
• Admin controls in workspace settings let you scope who can connect via MCP.

• Closed-source and provider-hosted, with no public GitHub repo, npm package, or self-hosted option.
• Public documentation on the exact tool schemas and server URL is sparse; most setup happens behind a Metaview login.
• Requires a Metaview subscription and a workspace admin to enable the integration; useful only if your team already runs interviews through Metaview.

• Zapier MCP for Metaview (https://zapier.com/mcp/metaview): exposes a narrower set of conversation actions via Zapier's MCP gateway.
• Greenhouse, Ashby, or Workable MCP integrations: if you need ATS-native data rather than interview intelligence, several ATS vendors and community projects expose MCP servers.
• Gong or Fireflies MCP-style integrations: alternative conversation intelligence platforms that surface call transcripts to LLMs, though oriented toward sales rather than recruiting.