Skip to main content
Connect Claude, Cursor, ChatGPT or Gemini to your Manglai platform and query carbon, water and waste data in natural language.

​
What Can You Do?

Inventory

Facilities and fleet vehicles. Create or update with confirmation.

Consumptions

Invoices, expenses, business travel, purchases and consumption records

Emissions

Carbon totals, line items, emission factors and compensation projects

Surveys

Commuting and materiality surveys and submitted answers

Reports

Saved legal reports (GEI, ISO 14064, GLEC, water, product PCF)

Custom metrics

Custom metric definitions and measurements

Alerts

Data-quality alerts for invoices, travel, vehicles, surveys and factors

Integrations

Bulk-import schemas, connectors and import history

Product footprint

BOM products and materials for product carbon footprint
Full tool list: Available tools.

​
How It Works

​
Security

  • Data tools — Read company data in the token scope
  • Confirmed writes — Action tools (upsert_building, upsert_vehicle, create_invoice, …) require userConfirmed: true after explicit user approval
  • Token-scoped access — Access is limited to what your token can see
  • No data storage — Data flows through directly, nothing is cached

​
What is MCP?

The Model Context Protocol connects your AI assistant to Manglai. Server resource: manglai://api/what-is-mcp.

​
Create or update inventory

Call with userConfirmed: false first (preview), then userConfirmed: true after approval. Resolve ids with list_buildings or list_vehicles. Agent guides: manglai://api/mutations/buildings and manglai://api/mutations/vehicles.

​
Architecture

​
Getting Started

Quickstart

Set up the Manglai MCP in 5 minutes