AgentBrains RAG icon

AgentBrains RAG

Retrieve information from AgentBrains RAG

Actions2

Retrieve Text

AI-generated

Summary

Searches a specified index for text documents matching a query and returns the results.

Inputs

  • index (required) — The ID or name of the knowledge base index to search. Defaults to 'general_helper_documents'. Can be selected from a list or specified via an expression.
  • query (required) — The text string to search for. When used as a tool, this is automatically filled by the AI model.
  • toolDescription — A description for the AI Agent to understand when to use this tool.
  • options — A collection of options including 'topK' (number of results to return, default 10) and 'metadata' (JSON object for filtering).

Output shape

The raw API response for the query, containing the matching documents and their metadata.

If an error occurs and 'Continue on Fail' is enabled, the item JSON contains an 'error' and 'details' field.

Examples

Example 1: Search for return policy terms in the default index

Index: general_helper_documents, Query: 'What are the return policy terms?'

Example 2: Retrieve top 5 results from a specific index with metadata filtering

Index: 'specific_id', Query: 'customer support', Options: topK=5, metadata={'category': 'support'}

Example 3: Use the node as a tool within an AI Agent workflow

Tool Description: 'Find product information', Query: [automatically filled by AI]

Links

Discussion