Spendflo MCP (Model Context Protocol)

Overview

Spendflo MCP (Model Context Protocol) is a remote server that connects MCP-compatible AI clients directly to your live Spendflo data. It enables you to query contracts, renewals, vendor spend, and procurement requests using natural language - without needing to navigate dashboards.

Key Capabilities

Spendflo MCP allows you to interact with procurement data conversationally:

Capability	What it Enables	Example Query
Renewal Tracking	Identify upcoming contract renewals and associated spend	“Which contracts renew in the next 60 days and what am I paying?”
Vendor Lookup	Retrieve contract details for specific vendors	“When does our Salesforce contract renew and what's the total cost?”
Spend Visibility	Compare actual vs contracted spend across vendors	“Show actual vs contracted spend for Q1 across our top vendors.”
Approval Bottlenecks	Detect procurement requests pending approval	“What procurement requests are stuck waiting on approvals right now?”
Portfolio Summary	Get a high-level overview of contracts and spend	“How many active software contracts do we have and what's our total annualized spend?”
Renewal Briefs	Generate summaries for upcoming renewals	“Draft a renewal brief for our top 5 contracts expiring this quarter.”

How to Connect Spendflo MCP

Follow the steps below to connect your AI client to Spendflo MCP:

Setup Steps

Step	Action	Instructions
Step 1: Open Connectors	Access connector settings in your AI client	Go to Settings → Connectors → Add Custom Connector
Step 2: Enter MCP URL	Add the Spendflo MCP endpoint	Enter: `https://api-v2.spendflo.com/mcp`
Step 3: Authenticate	Complete login via OAuth	Click Connect and sign in via the browser (no API keys required)
Step 4: Start Querying	Begin using natural language queries	Spendflo will appear as an active connector

Authentication Notes

Spendflo MCP uses OAuth-based authentication
No API keys or tokens are required
Ensure:
- Your browser session is logged into the same account as your AI client
- If not, sign out and retry the connection process

Supported AI Clients

Spendflo MCP works with any client that supports remote MCP over HTTP/SSE.

Compatible Clients Include:

Claude Desktop
Claude Web
Cursor
Windsurf
VS Code (Copilot-enabled)
Claude Code

How Spendflo MCP Works

You connect your AI client to Spendflo MCP
You enter a query in natural language
MCP interprets the query and retrieves relevant data
Spendflo returns live procurement insights
Results are displayed conversationally within your AI client

Best Practices

Use clear and specific queries (include vendor names, timeframes, or metrics)
Combine dimensions like time + vendor + spend for deeper insights
Use MCP for:
- Quick answers instead of navigating dashboards
- Generating summaries and reports
- Identifying operational bottlenecks

Example Use Cases

Use Case	Outcome
Procurement Reviews	Quickly assess renewals and vendor spend
Finance Planning	Compare contracted vs actual spend
Vendor Management	Track contract timelines and costs
Operations	Identify and resolve approval delays
Executive Reporting	Generate summaries instantly

Next Steps

Now that you have connected Spendflo MCP to your AI client, you can start querying your procurement data in natural language to track contracts, monitor renewals, and analyze vendor spend in real time. For more information on managing your contracts and procurement workflows, refer to our guide on Understanding the Agreements Listing Table.