---
title: "Builders — Deploy your agent on Vasco"
description: "Developer-focused — connect to Vasco's MCP server, the data architecture, and why it works for GTM teams."
canonical: "https://vasco.app/builders"
---

# Deploy your agent anywhere. Ground it on Vasco.

Claude or your own stack — wherever your agent runs, it needs the same revenue-data substrate underneath: consistent metrics, an account timeline from lead to cash, your actual plan, and integrity guarantees. Vasco provides it through one MCP endpoint: **`https://mcp.vasco.app/mcp`**, exposing **58 tools** across three layers.

Works with any [MCP-compatible](https://modelcontextprotocol.io) client; the full tool reference lives in the [Vasco MCP docs](https://help.vasco.app/en/articles/11487839-vasco-mcp).

## Quick start (~30 seconds)

Add Vasco as an MCP server, authenticate over OAuth (reads auto-approved; writes confirmed), then ask:

```
claude mcp add-json vasco '{"type":"http","url":"https://mcp.vasco.app/mcp","oauth":{...}}'
```

A real query composes multiple layers — e.g. "what's the win rate for expansion this quarter, and which at-risk accounts should we prioritize?" returns the metric (`expansion_win_rate` 0.42, n=12, +4pt QoQ), the plan gap ($82.5k of a $90k target booked), and the at-risk accounts (champion-change in the last 60 days), each with citations. Requires an authenticated Vasco user on a trial or paid workspace with sources connected; onboarding takes ~10 minutes and Vasco wires reconciliation for you. Supported clients include Claude Code, Claude Desktop, Cursor, a custom TypeScript SDK client, and raw `curl`.

## Why ground on Vasco: six categories of work you'd otherwise build

Between your agent and a real answer sit six engineering problems — each its own project, **~9 months total for a v1**. Vasco ships all of it on day one, behind one endpoint:

1. **Identity resolution** (6–8 weeks to build yourself) — `Acme Corp` is `account_8842` in Salesforce, `cus_K9aBz` in Stripe, `workspace_117` in your product, plus 47 contacts across Gong, Slack, and Outreach. Without resolution your agent reasons over seven strangers, not one company.
2. **Account timeline** (10–14 weeks) — reconstruct lead-to-cash across CRM, calls, tickets, threads, and contracts, with conversations placed in stage.
3. **Metrics definition** (4–6 weeks) — one locked, versioned definition per metric.
4. **Planning** (3–5 weeks) — quotas, capacity, coverage, forecasts to reason against.
5. **Memory** (8–12 weeks) — outcomes remembered across time.
6. **Causality** (a research project in its own right) — the *why* behind a number, not just the value.

## How the data flows

Sources in → reconciled → layered → served. Clients call `/mcp`; `/mcp` reads the three layers; the layers read a unified store; the unified store is the only thing that touches your sources. **Provenance:** every tool result carries source IDs and a citation, so you can drill from any number back to the row that produced it — even after reconciliation.

## The three layers, exposed as MCP

Every tool call falls into one of three composable layers:

- **Foundation** — *"What is the business? How does it perform?"* GTM Model (stages, channels, motions, functions, employees, dimensions, journeys), Source Mappings (CRM → GTM entity mapping, preview-before-save, versioned), Integrity Radar (data-quality score, issues, auto-reconciliation), and the Metric Engine (deterministic actuals, trends, conversions, breakdowns).
- **Plan** — *"Where is the org headed? How is it tracking?"* Goals & Scenarios (targets, what-if plans, apply & rollback), Plan Inputs (conversion rates, quotas, costs, retention, channel spend), Forecasts (projected revenue/pipeline/customers by month and motion), and Benchmarks (industry comps at P25/P50/P75).
- **Context** — *"Why do the numbers look the way they do?"* the Context Graph (entities, relationships, account context — traverse to find at-risk accounts, segment, or explain churn), Graph Schema (the intended first call before any traversal), Saved Queries (named, reusable, citable graph slices), and Artifacts (MetricQuery, ContextQuery, Reports, Slides — persisted, citable evidence).

## Why it works for GTM teams

The same data layer, governed by RevOps. Builders connect once; then every rep, every agent, and every seat reasons against the same definitions, plan, and graph. RevOps governs what agents see — without re-platforming the org. Full tool reference (name, signature, scope, response shape) lives in the docs.

## Get started

Connect in ~30 seconds using the quick start above (requires a Vasco workspace — trial or paid — with sources connected). [Browse all 58 tools in the docs](https://help.vasco.app/en/articles/11487839-vasco-mcp).