Setup
2
Add an AI Agent to your workflow
Create a workflow with an AI Agent node (for example, triggered by When chat message received) and attach a chat model to it.
3
Attach the MCP Client Tool node
Add the MCP Client Tool node as a tool of the AI Agent and configure it:
- Server Transport: HTTP Streamable
- Endpoint:
https://mcp.xpoz.ai/mcp - Authentication: Bearer Auth, with your Xpoz access key as the token
- Tools to Include: All (or select specific tools; see the tip below)
4
Run it
Ask the agent something that needs social media data, for example: “What are people saying about our brand on Reddit this week?” The agent discovers the Xpoz tools automatically and calls them as needed.
Example workflows
- Brand monitoring digest: Schedule Trigger → AI Agent (“summarize the last day of posts mentioning our brand across Twitter and Reddit”) with the Xpoz MCP tool → email or Slack node with the summary.
- Lead research: Chat Trigger → AI Agent that looks up a prospect’s Twitter/Instagram profile and recent posts before a sales call.
- Competitor tracking: Schedule Trigger → AI Agent using tracked keywords (
getTrackedItems,addTrackedItems) to maintain continuous coverage, then post changes to a channel.
Notes
- The MCP Client Tool node requires a recent n8n version; the HTTP Streamable transport is the current recommended option (SSE is legacy).
- Long-running Xpoz operations return an
operationId; agents can poll withcheckOperationStatus. For most searches the default fast mode returns results directly. - Each tool call consumes Xpoz credits per your plan.
Next Steps
MCP Tools
Browse all 48 tools across 4 platforms
Query Syntax
Get better search results with boolean operators and filters

