Skip to main content
Notion is the most connected server across Rippit customers, and the reason is usually the same: it is where the analysis is supposed to end up.

What the agent gets

  • Workspace search, including sources Notion itself has connected such as Slack, Google Drive, GitHub, and Jira.
  • Page and database reads, with SQL-style querying over data sources.
  • Page and database writes, if you enable them: create pages, update content and properties, add comments, create databases and views.
  • People and teams, for resolving who owns what.
Typical uses: writing a recurring insights digest into a Notion page, checking whether a rising issue is already documented, or updating a tracker database from what the agent found in the conversations.

Setup

Click Connect and sign in to Notion. Nothing to create first. Notion’s hosted MCP server is OAuth only. It does not accept integration or bearer tokens, so there is no application-identity option.

Identity and scope

The connection acts as the Notion user who authorized it and can reach exactly what that person can reach, private pages included. This matters more for Notion than for most connectors, because personal Notion workspaces tend to hold private notes alongside team content. If you are sharing the connection organization-wide, authorize with an account whose Notion access you are comfortable exposing to everyone.

Tools

Notion exposes more than forty tools. The read set is enabled by default.

Read

Write

Custom Agent sessions

Notion’s own Custom Agents can be started and followed from Rippit. notion-search-agents finds the agent; the tools below run it and read what it produced.

Plan prompts

Notion gates a few capabilities behind its own Business and Enterprise plans. When one of them is out of reach, the server offers a card explaining what the workspace would need. These tools only render that card; they do not read or write anything.