Learning Timeline
Key Insights
Overcoming Standard Tool Module Limitations
Standard tool modules often do not allow an AI Agent to determine the 'Database Item ID' dynamically. By using a Scenario as a tool, you give the Agent the freedom to choose exactly which row needs to be updated.
The Advantage of Dynamic Collection Output
Use the 'Dynamic Collection' type for the scenario output. This allows the AI Agent to receive all data (the bundle) from Notion once the update is complete, providing richer context for the Agent's next steps.
Mandatory Step: Activate Scenario
Don't forget to click 'Save' and 'Turn On' your scenario. If the scenario is 'Off', the AI Agent will be unable to call the tool and the process will fail.
Step by Step
Building a Scenario as a Notion Update Tool
- Open Make (formerly Integromat) and choose to create a new scenario to act as a Tool.
- Give the scenario a descriptive name, such as 'Update a Task'.
- Define the 'Input Structure' for the AI Agent to fill: Add a 'Database Item ID' (Text) field to identify which task needs to be updated.
- Add additional input fields such as 'Task Title' (Text), 'Due Date' (Date), and 'Status' (Select).
- For the 'Status' (Select) field, manually enter the options according to your Notion database (e.g., Backlog, Open, Done).
- Set the 'Output Structure' by adding an 'Updated Task' field and selecting 'Dynamic Collection' as the type.
- Click the button to enter the Scenario Builder.
- In the Notion 'Update a Database Item' module, select the correct database.
- In the 'Database Item ID' field, do not select a static item. Instead, map the data from the scenario input created earlier.
- Map other fields like Name, Status, and Due Date using dynamic input from the AI Agent.
- In the 'Return Output' module, select the output from the previous Notion module to send the full data (bundle) back to the AI Agent.
- Click 'Save' on the scenario.
- Important: Toggle the scenario switch to 'ON' to activate this tool for the AI Agent's use.
