Skip to main content
The VeluraHQ MCP connector links your workspace to Claude through the Model Context Protocol. Once connected, you can query your AI Search Analytics, trigger workflow runs, inspect Power Sheet status, and update Brand Kit rules without switching to the VeluraHQ UI — Claude becomes a natural-language interface to your entire workspace. Ask a question, get live data back, and act on it without breaking your train of thought.

​
What you can do from Claude

Query AI Visibility

Ask how your brand appears across prompts, platforms, and competitors. Get answers drawn from your live workspace data — mentions, citations, share of voice, and sentiment.

Trigger Workflows

Start research, refresh, or publish runs by describing the task in plain language. Claude translates your instruction into a workflow run with the right inputs.

Manage Sheets

Inspect row status, kick off bulk runs, and review outputs conversationally — without opening the Sheets interface.

Update Brand Kit

Refine voice rules and positioning directly from chat when your strategy shifts. Changes take effect immediately across all workflows that use Brand Kit.

​
Setting up MCP

1

Open MCP settings

Go to Settings > Integrations > MCP in your VeluraHQ workspace.
2

Generate your MCP token

Click Generate MCP Token. This token scopes the connection to your specific workspace. Store it securely — it has access to your workspace data and actions.
3

Add the server in Claude

In Claude, open Settings > Integrations (or the MCP configuration panel, depending on your Claude client). Add a new MCP server connection.
4

Enter the server URL and token

Paste the VeluraHQ MCP server URL and the token generated in step 2.
5

Test the connection

Verify the setup is working by asking Claude: “What is my current AI search visibility score?” Claude should return live data from your workspace.

​
Available MCP tools

VeluraHQ exposes 12+ tools through the MCP connector. Claude can call any of these on your behalf during a conversation:
  • query_ai_visibility — Get current mentions, position, citations, and share of voice
  • list_actions — Pull the prioritized Actions queue
  • trigger_workflow — Start a named workflow with input parameters
  • get_workflow_status — Check the status of a running workflow
  • list_sheets — List available Power Sheets and their row counts
  • run_sheet_row — Trigger a workflow run for a specific row
  • get_page_data — Fetch performance data for a URL (clicks, impressions, citations)
  • list_pages_by_filter — List pages matching a performance filter (e.g., decaying traffic)
  • get_brand_kit — Read current voice rules and Brand Kit settings
  • update_brand_kit_rule — Update a specific voice or tone rule
  • list_competitors — List tracked competitors and their visibility metrics
  • get_competitor_comparison — Compare your visibility against a named competitor

​
Example conversation

Here’s what a typical MCP-powered exchange looks like in Claude:
From there, you can ask Claude to trigger the workflow, check its status, or pull the draft output — all without leaving the conversation.

​
Access scope

The MCP connector only accesses data and actions you have explicitly authorized. You can restrict access per module — for example, analytics read-only with no publishing permissions — in Settings > Integrations > MCP > Scope. Adjust scope at any time without regenerating your token.
MCP setup takes less than 10 minutes from start to first query. Each workspace requires its own MCP token — if you manage multiple VeluraHQ workspaces (e.g., for different clients or brands), generate a separate token for each one and add each as a distinct MCP server in Claude.