Retainr icon

Retainr

Store, search, and retrieve AI agent memories.

Actions6

Memory → Search

AI-generated

Summary

Search memories using a natural language query, optionally filtering by namespace, tags, and similarity threshold, and limiting the number of results.

Inputs

  • main (required) — The data to pass through; each incoming item triggers a search. Input data is not consumed by the operation itself but can be referenced in parameter expressions.

Output shape

main

Each output item contains a json property with an array of matching memory objects (id, content, namespace, tags, score, etc.). If no matches, the array is empty.

Examples

Example 1: Find a user's stored preferences

Set Query to 'What are this user's preferences?', Namespace to 'customer:alice', Limit to 10, Threshold to 0.6, and optionally Tags to 'preference,settings'.

Links

Discussion