Vibe Prospecting MCP Server
Explorium's MCP server for B2B prospecting, contact discovery, and company enrichment driven by natural language queries.
Vibe Prospecting is Explorium's MCP server for B2B prospecting. It powers AI chat clients with live company and contact data so users can describe an ideal customer profile in plain English and have the agent build lead lists, research accounts, and prepare personalized outreach. The official repo (explorium-ai/vibeprospecting-mcp) ships as a Gemini CLI extension that points at the hosted MCP endpoint at https://mcp-gemini.vibeprospecting.ai/mcp, with OAuth handled automatically on first use.
The underlying dataset spans Explorium's network of 150M+ companies and 800M+ professionals aggregated from 50+ data sources, with firmographics, technographics, hiring trends, intent signals, funding events, and verified contact info. The server exposes company discovery, prospect discovery, enrichment, autocomplete for filter values, market statistics, and CSV export. It uses a lazy-evaluation workflow: the agent returns a small sample (5-10 rows), the user confirms, and only then is the full dataset processed and streamed into a session table for follow-up operations.
It is best suited for GTM, sales, and RevOps teams who want repeatable, data-intensive prospecting workflows inside an AI chat client rather than a traditional sales-intelligence UI.
Tools
| Tool | Description |
|---|---|
Company Discovery & Search |
Find companies by name, domain, industry, size, location, technographics, funding stage, or other firmographic attributes. |
Prospect Discovery |
Locate professionals at target companies by role, title, seniority, department, or location. |
Company Enrichment |
Enrich a company record with firmographics, technographics, financials, competitors, workforce trends, and recent events. |
Prospect Enrichment |
Enrich a contact record with verified email, profile data, career events, and social activity. |
Autocomplete |
Suggest valid filter values for industries, technologies, job titles, and locations to keep queries on schema. |
Market Statistics |
Return counts and aggregate stats for a given filter set before committing to full enrichment. |
CSV Match & Enrichment |
Match uploaded raw lead lists against the Explorium graph and enrich them with the requested attributes. |
Export to CSV |
Export the current session result table to CSV after the user confirms the sampled preview. |
Prerequisites
- A Vibe Prospecting / Explorium account (OAuth login is triggered on first use).
- An MCP-capable client. The official extension targets Gemini CLI; the hosted server URL can be used from any client that supports remote MCP servers.
Install via Gemini CLI (official path)
gemini extensions install https://github.com/explorium-ai/vibeprospecting-mcp
For local development:
gemini extensions link /path/to/vibeprospecting-mcp
On first invocation Gemini CLI opens an OAuth flow against vibeprospecting.ai and stores credentials automatically.
Use the hosted MCP endpoint directly
The extension is a thin wrapper around the hosted remote MCP server. Clients that support remote MCP servers can connect to:
https://mcp-gemini.vibeprospecting.ai/mcp
Example MCP client configuration:
{
"mcpServers": {
"vibe-prospecting": {
"url": "https://mcp-gemini.vibeprospecting.ai/mcp"
}
}
}
Explorium also publishes companion repos: explorium-ai/vibeprospecting-plugin (general plugin) and explorium-ai/vibeprospecting-cursor-plugin (Cursor). See the Explorium docs for the latest client-specific setup.
- Build a targeted lead list from a natural language ICP description, for example mid-market SaaS companies in the US that recently raised Series B.
- Find decision-makers (VP Engineering, Head of Data) at a set of target accounts and pull verified contact info for outreach.
- Enrich an existing CSV of company names or domains with firmographics, technographics, and headcount before importing to a CRM.
- Research a single account in depth: competitors, funding events, hiring trends, and tech stack, then draft personalized outreach from the same agent session.
- Size a market segment with live statistics (how many companies match) before deciding whether to run full enrichment and export.
- "Find 200 Series A or B fintech companies in the EU with 50-500 employees and pull the CTO and Head of Engineering for each."
- "Enrich this CSV of 500 domains with industry, employee count, tech stack, and last funding round."
- "Show me companies similar to Acme Corp that are hiring data engineers right now."
- "Pull verified work emails for the VP of Marketing at each company in my current list and export to CSV."
- "Research Snowflake: top competitors, recent funding, leadership changes, and current hiring trends."
- Official server from Explorium backed by a large B2B dataset (150M+ companies, 800M+ professionals across 50+ sources).
- Natural language workflow with a lazy preview pattern that avoids burning credits on bad queries before the user confirms.
- Covers the full prospecting loop in one server: discovery, enrichment, CSV match, and export.
- OAuth is handled automatically on first use, no API key management.
- The public GitHub repo is essentially a Gemini CLI extension wrapper, with limited install instructions for Claude Desktop, Cursor, and other MCP clients in the README itself.
- Individual tool names and parameter schemas are not enumerated in the README; capabilities are described at the workflow level rather than as a formal tools table.
- Requires a Vibe Prospecting / Explorium account, and full enrichment volume is gated by the account's plan and credits.
- Apollo.io MCP style community servers for B2B contact and company data.
- Clearbit (now part of HubSpot) for company and person enrichment via API.
- People Data Labs for programmatic person and company datasets.