Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.databar.ai/llms.txt

Use this file to discover all available pages before exploring further.

AI prompts let you run custom instructions against your table data, either as standalone enrichments or as the driving logic behind the AI Researcher. Databar provides tools to write prompts manually, generate them automatically, and save them as reusable templates for your entire workspace.
AI prompt editor with column references and a generated prompt

​
AI Prompt enrichments

An AI Prompt enrichment runs a custom prompt on every row in your table. Reference column values using the { syntax so the prompt adapts to each row’s data. Example: You have a table of company names and descriptions. Add an AI Prompt enrichment with the instruction: 
Classify {Company Description} into one of: SaaS, Marketplace, Hardware, Services.
For each row, the AI reads the description and writes the classification into a new column.
1

Open the enrichment panel

Click Enrich in the table toolbar and search for AI Prompt.
2

Write or generate your prompt

Enter your instruction in the prompt editor. Use { to insert column references wherever you need row-specific data.
3

Configure output columns

Select which response fields to add to your table.
4

Run the enrichment

Click Run to process rows. Start with a single row to validate the output.

​
AI Prompt Generator

Not sure how to phrase your prompt? Describe what you want in plain English and Databar will generate a well-structured prompt for you. The generator works for both AI Prompt enrichments and AI Researcher agents.
AI Prompt Generator interface with a natural-language description and generated output
1

Open the generator

In the prompt editor, click Generate prompt.
2

Describe your goal

Write a short description of what you want the AI to do — for example, “Summarize each company’s value proposition in one sentence.”
3

Review and edit

The generator produces a prompt with proper structure and column references. Edit it further if needed, then apply it to your enrichment.
The prompt generator is especially helpful when you need structured output formats (lists, JSON, specific fields) — it handles the formatting instructions for you.

​
AI Prompt Templates

Save any prompt as a workspace-wide template so you and your team can reuse it across tables without rewriting it each time. Templates are ideal for prompts you run regularly — lead qualification, content classification, data extraction patterns, and more.

​
Saving a template

After writing or generating a prompt, click Save as template, give it a name, and it becomes available across your entire workspace.

​
Using a template

When setting up an AI Prompt enrichment or AI Researcher, click Templates to browse your saved prompts. Select one, and the prompt editor is pre-filled with the template content. Adjust column references as needed for the current table.
Template browser showing saved prompt templates

​
Required vs. optional column references

When your prompt includes column references, each reference can be toggled between required and optional:
SettingBehavior
RequiredIf the referenced column is empty for a row, the enrichment skips that row. No credits are consumed.
OptionalThe enrichment runs even if the value is empty. The AI receives a blank value for that reference.
Use the toggle next to each column reference in the prompt editor to set its requirement level.
Leaving all references as optional may cause the AI to produce low-quality or irrelevant outputs when key data is missing. Mark the columns that are essential to your prompt as required.

​
Next steps

AI Researcher

Use AI agents to visit websites and extract structured data.

Enrichments

Learn how to add and manage enrichments on your tables.

Credits and billing

Understand how AI prompt runs are billed.