Action Node and Tools¶
One node with two faces. In the node list it is Redis Agent Memory, an action node the workflow runs to manage what the memory sub-node writes. In an AI Agent's Tool picker n8n lists the same node as Redis Agent Memory Tool, a tool sub-node the model calls when it decides to search or save a memory. Each operation is one REST call on the credential's store.
Resources and operations¶
| Resource | Operations |
|---|---|
| Long-Term Memory | Search (by meaning, for one user), Get Many (by filters), Create, Get, Get Sources (the sessions or documents a memory came from), Update, Delete, Move (to another namespace) |
| Session | Get, Get Many, Delete, Add Event, Delete Event |
| Namespace | Create, Get, Get Many, Update, Delete |
| Summary View | Get Result (one group's summary), Refresh, Create, Get, Get Many, Update, Delete |
Every field of every operation is listed in the Operations reference.
Using the node as a tool¶
Attach one Redis Agent Memory Tool per operation to the Agent's Tool input and pin everything but the content: set Resource, Operation and User ID yourself, and let the model fill Query (Search) or Text (Create) with Let the model define this parameter. Two instances cover most agents:
- Search memories: Long-Term Memory → Search, User ID fixed. The agent calls it when it needs something it was told before.
- Save memory: Long-Term Memory → Create, User ID fixed. The agent calls it to remember a fact right away, without waiting for extraction.
Never let the model fill User ID
The User ID is what keeps one user's memories away from another. Set it from the workflow, for example {{ $('Identify the user').item.json.userId }}.
Extraction is asynchronous, so a fact said in this turn may not be searchable until the next one; Save memory is how an agent makes a fact available at once.
Other safe tools: Get Sources and Search with Include Sources (read-only, lets the agent cite where it learned something) and Summary View → Get Result with the view and the User ID group pinned ("what do you know about me?" in one call).
Search and Get Many¶
Search ranks a user's memories by meaning against a Query and returns up to Limit results (1–100). Options narrow it: Memory Type, Namespace ID, Session ID, Topics (any of a comma-separated list) and a Similarity Threshold between 0 and 1 (0 returns the closest matches regardless of score).
Get Many lists memories by filters alone, without a query: User ID, Memory Type, Namespace ID, Session ID, Topics. At least one filter is required. It pages through the server's results with Return All or Limit. Use it for clean-up ("every memory from this session") and audits; use Search when a question is involved.
Create, Update, Delete and Move¶
- Create stores a Text for a User ID, with optional Memory ID, Memory Type, Namespace ID, Session ID and Topics. Creating the same Memory ID again replaces the memory, which is how a workflow keeps a single fact current.
- Update changes any of the same fields on an existing memory. Topics replaces the existing topics.
- Delete removes one memory. Deletion is permanent.
- Move places a memory in another namespace, chosen from the store's list or given as an ID or expression.
Output¶
Every operation that returns records has a Simplify switch, on by default, which keeps the fields a workflow usually needs:
| Record | Simplified fields |
|---|---|
| Memory | memoryId, text, ownerId, memoryType, topics, namespaceId, sessionId, createdAt; with Include Sources on (Search, Get, Get Many), sources: sourceId, kind, sessionId or documentId, eventIds |
| Source (Get Sources) | sourceId, kind, sessionId, documentId, revision, eventIds (every event behind the source, in one list), evidenceCount (the number of evidence entries, not of events), capturedAt (when the first evidence entry was recorded) |
| Session (Get) | sessionId, ownerId, namespaceId, the summary text and the events, each with eventId, actorId, role, text, createdAt |
| Session (Get Many) | sessionId, ownerId, namespaceId, createdAt, updatedAt |
| Namespace | namespaceId, name, scope, state, parentId, ownerId |
| Summary view | viewId, name, visibility, groupBy, state, outputFormat, refreshMode, updatedAt |
| Summary view result | viewId, status, stale, the text or data, computedAt, sourceMemoryCount, summarizedMemoryCount, truncated and, when present, sourceMemoryIds and errorCode |
Switch Simplify off for the API's object as it came. Delete operations return { deleted: true } with the ID that was deleted.
Resource locators. The Namespace field of Namespace → Get, Update, Delete and Long-Term Memory → Move offers the store's namespaces from a list, or takes an ID or expression; the View field of Summary View does the same for the store's views.
Errors¶
The node turns the API's Problem Details into n8n errors: the server's detail becomes the message and title (HTTP n) the description. A 404 with the resource-not-found type is the only "not found" the node recognises; a read-only API key answers 403. With Continue On Fail on, the error lands on the item and the workflow can branch on it.