Smart Memory

Stores, retrieves, and manages conversational history for AI and chatbot applications. Supports multiple backends and flexible data structures.

Actions3

Insert Message

AI-generated

Summary

Insert a message into memory for the specified session, appending it by default or replacing the session’s existing memory. Choose a storage backend and configure message metadata, memory limits, and output formatting as needed.

Inputs

  • Session ID (required) — Identifies the conversation session to update.
  • Message Content (required) — Text to add to the session’s memory.
  • Insert Mode — Append the message to existing memory (default), or replace the memory with this message.
  • Storage Backend — Select in-memory, Redis, PostgreSQL, or SQLite storage. In-memory storage is lost on restart; Redis supports a TTL, and PostgreSQL or SQLite allow a table name, with SQLite also accepting a database file path.
  • Memory limits — Set the memory window size, pruning step size, and maximum stored message count to control retained history.
  • Write to Storage — Write immediately (default), or disable writes to buffer them for a later insert with writing enabled.
  • Metadata — Optionally attach identity, chat, message, and reply context fields to the inserted message.
  • Output Options — Choose structured output or LLM chat messages, and configure which timestamps, chat details, identity fields, and reply context to include.
  • Log Level — Sets the verbosity of node logging.

Output shape

A single item containing the updated memory.

The output format and included fields depend on Output Options. The default format is structured; LLM Chat Messages returns a role/content message list.

Links

Discussion