Skip to main content
The Model Context Protocol (MCP) is an open protocol for connecting AI assistants (Claude, ChatGPT, Cursor, VS Code Copilot, and others) to external tools and data over a standard JSON-RPC interface. Instead of copy-pasting into a dashboard, you talk to Scanova in plain language from inside the AI tool you already use — “list my active QR Codes,” “create a QR Code for our new landing page,” “show me last week’s scan analytics.” Scanova runs a dedicated MCP server for this, rather than an API wrapper glued onto an existing endpoint. The server is open source: github.com/trycon/scanova-mcp.

​
What you can do with it

Once connected, your AI assistant can act on your Scanova account through the same Management API the dashboard itself uses. Broadly, that covers:
  • QR Codes — create, list, update, retrieve, activate, deactivate, delete, download as PNG, JPG or PDF, tag, and restore from the trash, plus a health check for codes that need attention
  • Design — preview designs with scan-safety checks and a real scan test before anything is saved, then apply the one you choose
  • Categories and validation — look up what a category needs, and check a payload before creating anything with it
  • Analytics — account-level usage counters, a scan overview, per-QR Code scan analytics by device, geography, and date, and reports generated and emailed to you
  • Folders and tags — create, list, rename, and delete folders, move QR Codes between them, and read or set the tags in use
  • Forms and lead lists — create forms from a title and questions, read their responses and analytics, set up email alerts for new responses, and attach forms or lead lists to a QR Code
  • GS1 recalls — post, change, or end a batch/lot recall notice for GS1 QR Codes
  • Users, roles, and activity — list, invite, and remove users, change their roles, create custom roles, and read who did what in the account
  • Your plan and billing — read the active plan, the plans you could move to, what a downgrade would affect, and your payment and order history
  • The rest of your workspace — custom domains, bulk operations, the media library, integrations and webhooks, Conversion Tracking sites, and landing page templates, read-only
  • Scanova’s own docs — a bridge tool lets the assistant search or read this documentation site live while it works, so it can look up something like the pattern-info schema before calling another tool
See Available tools for the complete, per-tool list with descriptions.

​
Where it lives

In the dashboard, go to Integrations and open the Model Context Protocol card, filed under Automation. This opens a setup panel with your server URL, per-client connection steps, and a token fallback for clients that don’t support OAuth — see Setup for the full walkthrough.

​
Access requirements

The MCP integration is gated behind an entitlement quota codename, INTEGRATION_MCP, checked on both the dashboard card (to show Connect instead of Upgrade) and on every request the MCP server makes to the Management API on your behalf. This is a plan-managed grant rather than a quota you turn on yourself.
INTEGRATION_MCP is distinct from the MANAGEMENT_API quota that the plain Management API uses, so holding one does not imply the other. If the Model Context Protocol card shows Upgrade instead of Connect, your account does not have it yet — contact support@scanova.io.

Setup

Generate a connection, and wire it into Claude, ChatGPT, Perplexity, or a config-file-based client.

Available tools

Every tool the server exposes, grouped by what it works on.
  • MCP setup — connect Claude, ChatGPT, Perplexity, or a config-file client to the server described here
  • Available tools — the complete, per-tool reference for everything the server exposes
  • Management API overview — the environment=mcp token type and its INTEGRATION_MCP quota gate, documented in full