We have Apay Sensei Supervisor MCP connector which is used for generating deep dives based on the datasets available in Apay Sensei Supervisor. I want to use this Apay Sensei Supervisor MCP connector with the automation which we created so that we can utilize the automation to do the deep dive, I wanted to check if there is a documentation on how to achieve this.
Hi @nitea and welcome to the community!
Happy to point you in the right direction, though I want to be upfront that Apay Sensei Supervisor is an internal Amazon connector, so there won’t be public community documentation specific to it. For anything tied to that connector itself, the connector’s owning team or your internal Quick support channel is the right place, since they own how it’s exposed and what it supports.
On the general side, one clarifying question so I point you at the right docs: when you say the automation you created, do you mean a Quick Flow or a Quick Automate (I assume an Automate, but want to be sure)? Those are the two different automation capabilities in Quick, and how you bring a connector’s tools into each one isn’t quite the same. If it’s a Flow, action steps are the piece that lets it call out to a connector as part of the run. If it’s Quick Automate, it’s handled through the actions/integrations available in that automation. Either way, the connector has to be available as an action inside the builder for you to call its deep-dive capability directly
But as for general documentation, I would give these a read:
- Work with integrations in Amazon Quick.
- Model Context Protocol (MCP) integration - Amazon Quick
- Action steps - Amazon Quick
- Building automations - Amazon Quick
Hope this helps!
Hello, Thanks for the response, I am using Quick Automate.
Hi @nitea ,
Quick Automate does support MCP connectors (the compatibility matrix lists MCP as supported for Automations: Supported connector types and available actions - Amazon Quick ), so this is doable. An MCP integration registers each tool the server exposes as an action you review and turn on, so once Apay Sensei Supervisor is connected, its deep-dive tool becomes an action you can drop into the automation and whose output you capture in a variable for later steps. The one gotcha is that integrations are scoped to the automation group your project lives in, so the connector has to be set up for that group first, which takes owner permissions on it. The MCP integration and building automations pages cover both halves: Model Context Protocol (MCP) integration - Amazon Quick and Building automations - Amazon Quick .
Hope this helps!