skip to content

Use Notion’s Property Description As Text → DB add-itor

Last Updated:
Coding Notion LLMs AI

Val town script here (which you can import as a module if you want!)

Demo using the template Money Database
A video demonstrating the use of function - with an http endpoint
And the outcome of the demo video above. Shows 3 entries, 2 of type debit, 1 of credit, across various inferred categories etc.
[Image version] And the outcome of the demo video above. Shows 3 entries, 2 of type debit, 1 of credit, across various inferred categories etc.
[Image version] And the outcome of the demo video above. Shows 3 entries, 2 of type debit, 1 of credit, across various inferred categories etc.
  • Uses instructor and open ai (with gpt-4-turbo) to process any content into a notion database entry.
  • Use addToNotion with any database id and content. 
await addToNotion(
  "DB_ID_GOES_HERE",
  "CONTENT_GOES HERE"//"for example: $43.28 ordered malai kofta and kadhi (doordash) [me and mom] jan 3 2024"
);

Prompts are created based on your database name, database description, property name, property type, property description, and if applicable, property options (and their descriptions).

Supports: checkbox, date, multi_select, number, rich_text, select, status, title, url, email. Filters properties where either type is title or description starts with ✨.

  • Uses NOTION_API_KEY, OPENAI_API_KEY stored in env variables and uses Valtown blob storage to store information about the database.
  • Use get_notion_db_info to use the stored blob if exists or create one, use get_and_save_notion_db_info to create a new blob (and replace an existing one if exists).