Message History component (AI Flowi Workflow)

Message History stores chat messages or retrieves messages already stored in AI Flowi Workflow tables or an external memory. Place it where a chat flow needs to save a message or supply earlier messages to a later step.

Message History is documented in the catalogue; not production-tested by Ai Flowi.

By Ai Flowi · Published

Where Message History fits

The 1.9.2 catalogue, exported 9 October 2026, lists Retrieve and Store modes. Decide which task this node serves before wiring it: saving the current message and retrieving earlier messages are separate configuration choices. Only order is marked required in this export; a field marked optional is not a promise that every mode works with it empty.

Related: Mem0 integration for AI Flowi Workflow.

Inputs

Message History input names, types and required flags below come from the component catalogue. For Message History, “Not listed in export” means no accepted-connection list was supplied; it is not a claim that every type is supported.

On small screens, scroll either Message History table sideways to read every column.

Message History inputs
NameTypeRequiredAccepted input typesPlain explanation
context_idstrNoMessageContext identifier setting. The export does not define its relationship to session_id.
memoryotherNoMemoryConnection for an external Memory object, rather than a Message or JSON record.
messagestrNoMessageChat message supplied to the history component; relevant when configuring a store step.
modetabNoNot listed in exportSelect Retrieve or Store for the history task.
n_messagesintNoNot listed in exportMessage-count setting. No default, maximum or selection rule is provided in the export.
orderstrYesNot listed in exportRequired ordering choice: Ascending or Descending.
senderstrNoMessageSender setting for the history operation; no matching rules are specified.
sender_namestrNoMessageSender-name setting. Check the stored message values before choosing it.
sender_typestrNoNot listed in exportSender category selection: Machine, User, or Machine and User.
session_idstrNoMessageSession identifier setting. Use a synthetic session value while checking retrieval.
templatestrNoMessageTemplate text setting. The catalogue does not publish its variable syntax.

Outputs

The Messages output is typed Message; the Table output is typed Table. Choose the output that the next component accepts. A Table connection is not interchangeable with a Message connection just because both concern the same conversation.

Message History outputs
NameDisplay nameType
messages_textMessagesMessage
dataframeTableTable

When to use Message History

Use it when the next step needs stored conversation content or when your design explicitly saves a chat message. Chat Input gets a current Playground chat input; Chat Output displays a message there. Message History addresses storage and retrieval instead.

When not to use Message History

Skip a history node when your flow only needs the current message. Prompt Template creates a prompt with dynamic variables, while Language Model runs a configured model. Neither purpose is a reason by itself to add a separate history step.

example configuration, not a recorded test

{
  "mode": "Retrieve",
  "order": "Ascending",
  "n_messages": 5,
  "session_id": "demo-session"
}

This sketch requests Retrieve mode with an Ascending order and a message-count setting of 5 for a synthetic session. It does not show a recorded retrieval result, create prior messages, or establish how the storage backend interprets the count.

Limitations

  • The catalogue supplies input and output types, not retention periods, storage capacity, persistence guarantees or deletion behaviour. Confirm those requirements before relying on stored history.
  • No isolation or access-control contract for context_id or session_id is documented in this export. An identifier setting alone is not evidence of authorization between users.
  • External memory configuration and availability need their own checks. The template syntax, empty-history result and interaction between sender settings are not specified by the catalogue.

Chat Input · Chat Output · Prompt Template · Agent · Language Model

Node runs and usage

Plans: Go from USD8/month billed yearly (USD10 monthly), Plus from USD16 (USD20 monthly), Pro 100 from USD80 (USD100 monthly), Pro 200 from USD160 (USD200 monthly) and Pro 500 from USD400 (USD500 monthly), with 5,000 to 250,000 Flowi Tokens/month and top-ups; 1 token per node run; third-party API and model usage is not included.

If the surrounding chat flow calls a model or third-party API, that usage is not included in the Workflow plan; count history node runs alongside the other executed steps. Review pricing for Message History.