I love Coda but I feel obligated to reluctantly use Notion due to its vast library of connectors. In an ideal world, Coda is the database, Claude can read/write to it, and Granola funnels all meeting notes into Coda. I can set this up in 5 minutes in Notion and it’s driving me crazy that I can’t here. I feel like Coda is going to lose a large base of users without this flexibility. Thoughts?
Hey @Jordan_Moore1 ! Just wrote you a DM re: MCP access.
Having a Coda MCP would be awesome… I use Claude and Cursor. So It would help me bring in data generated by the Ai and also read from Coda
@Edward_Battistini1 Just sent you a DM!
Are you aware of Coda‘s packs? There are 500+ available.
Packs help for fixed API connections, but they don’t really replace MCP once Claude or Cursor becomes the interface. The gap is bidirectional context: letting the model read a doc or table, write rows back, and chain that with other tools in one loop. I can patch pieces of this today with Packs + automations + webhooks, but it gets brittle fast once you want doc context, row updates, and meeting-note ingestion in the same workflow.
The OP was about Notion was necessary for the OP, because Notion has Connectors and CODA does not.
As far as MCP goes, the Coda/Superhuman/Grammarly MCP is in Beta at the moment.