In-Memory Checkpointer

In-Memory Checkpointer keeps the state of an agent’s conversation (the messages and the steps it has taken) in NappAI’s temporary memory. A checkpointer is what lets an agent pick up a conversation where it left off instead of starting from scratch with every message.

With this component, that state is not saved in any database: it lives only while the run lasts and is lost when it ends or if NappAI restarts.

Component in development

This component only appears in installations with development components turned on, and for now no component in the catalogue accepts its Checkpointer output. To make an agent remember conversations, turn on Use memory in the agent, which stores the history for you, or connect Redis Memory or Postgres Memory to its Memory input.

When to use it

• Tests where it does not matter that the conversation is lost when the run ends.
• Making sure no conversation data is stored outside the run.

Configuration

This component has no fields to configure.

Output

Checkpointer is the place where the conversation state is kept, in temporary memory. It sits at the top of the component.

Good practices

• Don’t use it when the conversation must last longer than one run: everything stored is lost when it ends. To keep it permanently you need a database, as in Postgres Checkpointer or MySql Checkpointer.

Related components

• Postgres Checkpointer
• MySql Checkpointer
• In-Memory Store
• Redis Memory