Text Input Node

The Text Input node emits a single record built from one or more named text fields you enter at design time. It is a source node — most useful as the starting point for automations that feed an LLM or Agent rather than a dataset.

When to Use

Use Text Input when your automation starts from fixed text — a brief, prompt, query, or set of named instructions — instead of an uploaded file or audience. For example, supply a brief field and a tone field that a downstream LLM node references.

Configuration

Example

Provide a research brief and tone as two fields:

fields:
  - name: brief
    value: "Research the top 5 competitors of Acme Corp and summarize their positioning."
  - name: tone
    value: "concise and analytical"

Output: a single record { brief: "Research the top 5…", tone: "concise and analytical" }.

Notes

• Emits exactly one record. Each field becomes a column named after the field's name.
• Field names must be unique and non-empty; duplicates fail validation.
• Downstream LLM and Agent nodes can reference field values via {{name}} in their prompts (see template syntax).
• The node also exposes a Combine input handle, which lets it merge upstream data into its single record when wired into a graph.