I Released a Hyros MCP Server — 100 Installs in 24 Hours

MCP — Model Context Protocol — is an open standard developed by Anthropic that lets AI assistants like Claude connect directly to external data sources and tools. Think of it as a universal adapter between AI models and the services you already use.

Instead of copying and pasting data into Claude or writing custom API wrappers every time you need to ask a question about your business, MCP lets the AI reach directly into your systems — read data, run queries, and take actions — all from a natural language conversation.

Claude Desktop, Cursor, Windsurf, and a growing list of AI tools all support MCP natively. You install an MCP server, point your AI tool at it, and suddenly your AI assistant has real-time access to your data.

Hyros holds some of the most valuable data in your entire marketing stack — which ads are driving real revenue, which leads converted, what your true ROAS is across every channel. But accessing that data has always required logging into the Hyros dashboard, building custom reports, or wiring up API calls.

With the MCP server running, you can ask Claude questions like:

Claude queries your live Hyros data and responds in plain English. No dashboard-switching. No exports. No waiting for reports to generate.

The MCP server wraps the core Hyros API endpoints as MCP tools — meaning Claude can call them on your behalf based on your prompts. Current tools included:

• get_leads — Query leads by date range, source, status, or email. Returns full lead objects with UTM data and attribution.
• get_revenue_events — Fetch revenue events with filters for amount, date, product, and originating lead.
• get_campaigns — List active and historical campaigns with performance metrics from your Hyros account.
• get_calls — Retrieve call logs with duration, outcome, and lead attribution data.
• create_lead — Create a new lead in Hyros with source, UTM data, contact info, and custom fields.
• track_revenue — Push a revenue event tied to an existing lead — amount, currency, product, and transaction ID.
• track_call — Log a call event against a lead for attribution tracking.
• get_attribution_report — Pull a cross-channel attribution summary for a given date range.

Installation takes about 5 minutes. You need Node.js 18+ and your Hyros API key.

{
  "mcpServers": {
    "hyros": {
      "command": "hyros-mcp",
      "env": {
        "HYROS_API_KEY": "your-hyros-api-key-here"
      }
    }
  }
}

I released hyros-mcp on March 25, 2026. Within 24 hours, the npm package hit 100 installs. That's not a huge number on its own — but for a niche tool targeting a specific marketing platform, it's a clear signal: people are actively looking for this.

The most common feedback I got: marketers want to ask questions about their attribution data in plain English without building dashboards or writing SQL. The MCP format is perfect for this because it meets them where they already are — in Claude or Cursor.

The second most common request: more read-heavy reporting tools. The current version covers the core CRUD operations and a basic attribution report. I'm extending it to cover deeper reporting endpoints — campaign comparisons, lead-to-close funnels, and channel-level ROAS summaries.