Vruum MCP server
{
"mcpServers": {
"vruum": {
"command": "npx",
"args": [
"-y",
"@vruum/mcp@latest"
],
"env": {
"VRUUM_MCP_TOKEN": "<vruum_mcp_token>",
"VRUUM_TOKEN": "<vruum_token>",
"VRUUM_MCP_URL": "https://..."
}
}
}
}{
"mcpServers": {
"vruum": {
"command": "npx",
"args": [
"-y",
"@vruum/mcp@latest"
],
"env": {
"VRUUM_MCP_TOKEN": "<vruum_mcp_token>",
"VRUUM_TOKEN": "<vruum_token>",
"VRUUM_MCP_URL": "https://..."
}
}
}
}{
"mcpServers": {
"vruum": {
"command": "npx",
"args": [
"-y",
"@vruum/mcp@latest"
],
"env": {
"VRUUM_MCP_TOKEN": "<vruum_mcp_token>",
"VRUUM_TOKEN": "<vruum_token>",
"VRUUM_MCP_URL": "https://..."
}
}
}
}{
"servers": {
"vruum": {
"command": "npx",
"args": [
"-y",
"@vruum/mcp@latest"
],
"env": {
"VRUUM_MCP_TOKEN": "${input:vruum_mcp_token}",
"VRUUM_TOKEN": "${input:vruum_token}",
"VRUUM_MCP_URL": "${input:vruum_mcp_url}"
}
}
},
"inputs": [
{
"id": "vruum_mcp_token",
"type": "promptString",
"description": "Vruum mcp token",
"password": true
},
{
"id": "vruum_token",
"type": "promptString",
"description": "Vruum token",
"password": true
},
{
"id": "vruum_mcp_url",
"type": "promptString",
"description": "Vruum mcp url",
"password": true
}
]
}{
"mcpServers": {
"vruum": {
"command": "npx",
"args": [
"-y",
"@vruum/mcp@latest"
],
"env": {
"VRUUM_MCP_TOKEN": "<vruum_mcp_token>",
"VRUUM_TOKEN": "<vruum_token>",
"VRUUM_MCP_URL": "https://..."
}
}
}
}file .mcp.json (in your project root)file claude_desktop_config.jsonfile ~/.cursor/mcp.jsonfile .vscode/mcp.jsonfile ~/.codeium/windsurf/mcp_config.jsoninstall npx -y @vruum/mcpversion 0.4.0
What it does
Vruum MCP gives agents access to an AI revenue platform with 29 tools for prospect research, outreach management, pipeline inspection, deal tracking, and revenue monitoring. It is for sales teams and revenue operators who want to automate research, campaigns, and CRM workflows through Claude or other AI agents.
- Research company fit and find warm introductions to prospects before outreach.
- Draft outreach messages, review them for approval, and manage email and LinkedIn campaigns.
- Inspect pipeline risk, track deal progress, and get daily briefings on pending actions and new replies.
Tools this server exposes
| Tool | Does |
|---|---|
| get_daily_briefing | Returns pending approvals, new replies, stalled deals, and a recommended next action for the day. |
| get_next_actions | Identifies the next steps in the revenue workflow. |
| inspect_pipeline | Surfaces the most at-risk deals in the pipeline. |
| search | Searches for prospects or accounts. |
| fetch | Retrieves details about a prospect or account. |
| research | Gathers website, funding, careers signals, and ICP match assessment for a company. |
| import_prospects | Imports a list of prospects. |
| find_warm_path | Identifies verified warm introductions and connector candidates to a person. |
| get_person_360 | Returns comprehensive information about a person. |
| manage_person | Updates or modifies person records. |
| get_outreach_review | Surfaces outreach drafts for approval, editing, or rejection. |
| manage_messages | Handles email and messaging operations. |
| manage_outreach | Manages outreach campaigns and their execution. |
| manage_relationship_action | Tracks relationship-building actions. |
| get_engagement_review | Reviews engagement and interaction records. |
| manage_engagements | Updates engagement tracking. |
| get_content_review | Reviews content drafts for approval. |
| manage_content | Writes, schedules, and publishes LinkedIn posts and content. |
| manage_campaign | Creates, updates, or manages outreach campaigns. |
| get_campaign_outcomes | Reports on campaign performance: contacted, replies, meetings booked, and cohort consistency. |
| get_performance_metrics | Returns performance and outcome metrics. |
| get_deal_360 | Returns comprehensive information about a deal. |
| manage_deal | Updates or progresses a deal through the sales cycle. |
| get_revenue | Accesses Stripe-backed revenue data and truth. |
| manage_revenue | Updates revenue records. |
Before you enable it
An MCP server runs with whatever access you give it. Start with a development account rather than production, and check which tools the server exposes before you let an agent call them unattended.