Connect CLS++ Memory to n8n

Give your n8n AI Agents persistent memory with the CLS++ Memory MCP server. Tools exposed: recall_memories, store_memory, who_am_i.

Quick install (MCP Client Tool node)

In any workflow that has an AI Agent node, add an MCP Client Tool sub-node to the agent's Tools input and configure it:

The node type is @n8n/n8n-nodes-langchain.mcpClientTool (typeVersion ≥ 1.2 for HTTP Streamable). SSE is deprecated — use HTTP Streamable.

Or import the ready-made workflow

Import examples/integrations/n8n-cls-memory-template.json via Workflows → Import from File. It wires a Chat Trigger → AI Agent with the CLS++ Memory MCP Client Tool attached.

OAuth login

The CLS++ endpoint uses OAuth 2.1 + PKCE with Dynamic Client Registration. Choose Authentication → MCP OAuth2 on the node, create the credential, and complete the browser sign-in n8n prompts you with. n8n then refreshes the token automatically.

Fallback: API key (Bearer auth)

Prefer a static token? Get an API key at https://www.clsplusplus.com, then on the MCP Client Tool node set:

• Authentication: Bearer Auth
• Credential → Bearer Token: your CLS_API_KEY

(The Endpoint stays https://www.clsplusplus.com/mcp.)

Verify it works

Open the workflow's chat and ask: "Use the CLS++ memory tools to store that my team uses n8n for automation, then recall what my team uses." The agent should invoke store_memory and then return "n8n for automation" via recall_memories. You can also confirm the node loaded the three tools in the MCP Client Tool node's Tools to Include dropdown.