Skip to content

MCP overview

Stackify supports the Model Context Protocol (MCP) — an open standard for connecting AI assistants and LLM-powered tools to data sources and services.

With MCP, you can give AI tools like Claude, Cursor, or custom agents direct, authorised access to your Stackify workspace — so they can read and write cards, fields, and records on your behalf.

How it works

Stackify acts as an MCP server. An AI tool (the MCP client) connects to Stackify using your authorisation, and can then:

  • List cards in your workspace
  • Read records from a card
  • Create or update records
  • Discover the field schema of any card

All access is gated by your workspace permissions — the AI tool can only do what your account is allowed to do.

Use cases

  • Ask an AI assistant to query your Stackify data in natural language: "Show me all projects due this week"
  • Let an AI agent write processed results back to a Stackify card
  • Build custom AI workflows that read from and write to structured datasets without writing API glue code

Authorisation model

MCP access uses OAuth 2.1 with PKCE. You explicitly authorise each AI tool that requests access. Authorisation can be revoked at any time from Settings → Connected apps.

See Connecting AI tools for setup instructions.