Connect agentiQu to Claude

Connector details

https://api.agentiqu.com/mcp/

Include the trailing slash: the URL without it redirects, and some clients do not follow the redirect.

Sign-in

OAuth 2.1. Clients identify themselves with a client ID metadata document (CIMD) or dynamic client registration. No API key needed - you sign in to agentiQu in your browser and approve the connection.

Workspace

On the consent screen you choose which agentiQu workspace the connection may use. Everything it does happens in that workspace, as you.

Transport

Streamable HTTP (a remote MCP server).

You need an agentiQu account to sign in.

Set up

Claude on the web and Claude Desktop

1. Open Customize → Connectors → Add and choose Add custom connector.
2. Name it agentiQu, paste the connector URL, and click Continue. Keep the detected defaults (Sign in now, and Claude's published identity as the OAuth client), then click Add.
3. Click Connect, sign in to agentiQu, choose a workspace, and click Allow.

Claude Desktop uses the connectors of the Claude account you are signed in to, so adding the connector on the web is enough.

Claude Team and Enterprise

An organization owner adds the connector once for the whole organization in the organization's connector settings, using the same URL. Each member then clicks Connect and signs in as themselves, so every action is attributed to the person who took it.

Claude Code

Add the connector from your terminal:

claude mcp add --transport http agentiqu https://api.agentiqu.com/mcp/

Then run /mcp inside Claude Code, select agentiqu, and choose Authenticate. Your browser opens the same approval screen. Tokens refresh automatically.

Other MCP clients

Any client that supports remote (HTTP) MCP servers can connect with the same URL. Clients that support MCP authorization go through the same browser sign-in; for clients that do not, use an API key as described below.

Use an API key instead

For CI, scripts, and clients without sign-in support, authenticate with a workspace API key. Create one in the agentiQu dashboard under Settings → API keys and send it in an X-API-Key header. The key is scoped to the workspace it was created in and can be revoked from the same page.

claude mcp add --transport http agentiqu https://api.agentiqu.com/mcp/ --header "X-API-Key: agt_xxxxxxxxxxxxxxxx"

What the connector can do

Once connected, describe what you want in plain language and Claude picks the tools. A connection acts as you, in the workspace you picked, with the same reach as a workspace API key.

Website simulations

Create, update, and delete simulations for websites and conversational bots; start and cancel runs; follow a run's progress and health; and read each synthetic user's step-by-step session and session-form answers.

Ad testing

Create, update, duplicate, and delete ad sets in your ad library and run marketing simulations against them. Read campaigns, daily KPIs, and funnel metrics synced from your ad platforms, track predicted against observed KPIs, and build marketing validation plans and reports.

Voice calls

List the calls in a run and read their details and transcripts, add call notes and set tags, look up saved voice use cases, and view voice schedules and their slots, including pausing or resuming a schedule.

Reports and insights

Read a run's aggregated insights, list its sessions, and export a run as JSON, YAML, or an agent-ready prompt. Browse your saved audiences and their personas.

Account and billing

Check which user and workspace the connection acts as, the plan catalog, your credit balance, monthly API quota usage, and spend limits. Upgrading a plan returns a Stripe Checkout link for you to open; nothing is charged unless you complete checkout.

Read-only tools and actions that need confirmation

Every tool is annotated for your client. Tools that only read data are marked read-only. Tools that delete or overwrite data, start a run, or upgrade your plan are also marked destructive, so your client can ask you to confirm each call instead of running it automatically.

Starting a run spends your workspace's credits. Connecting and listing tools costs nothing; each tool call is metered against your plan's monthly API request quota exactly as the equivalent REST API call would be.

Example prompts

Try one of these after connecting:

• Using agentiQu, create a website simulation for https://example.com with an audience of US small-business owners shopping for accounting software, run 5 journeys in read-only mode, and summarise where they dropped off.
• Show me my last three agentiQu runs and, for the most recent completed one, the top usability issues from its insights.
• List my synced ad campaigns from the last 30 days, find the funnel step with the worst drop-off, and suggest a simulation to diagnose it.
• What's my agentiQu credit balance and API quota, and would a 20-journey run fit within my spend limits?

Disconnect

To revoke access, open the agentiQu dashboard and go to Settings → Connected apps. A revoked connection stops working immediately. You can also remove the connector from your client's own settings. Sign-in connections that stay idle for more than 30 days expire and need to be approved again.

Data and privacy

The connector reads and writes data in the workspace you picked: simulations, runs, insights, sessions, calls, ad sets, and synced campaign data. Tool results are returned to your MCP client, such as Claude, and are handled there under that provider's terms.

Read our Privacy Policy and Terms of Use.

Questions about the connector? Email hello@agentiqu.com.