Introduction to Adzviser and OpenClaw
Adzviser integrates with OpenClaw via MCP (Model Context Protocol) to give you a powerful, open-source command-line AI agent for querying your marketing and sales data.
What is OpenClaw?
OpenClaw is an open-source AI agent you can run on your own computer or server. You choose the model and tools it connects to, including a cloud model provider or a supported locally hosted model. With OpenClaw, you can:
- Run AI-powered queries directly from your terminal
- Connect to multiple MCP servers simultaneously
- Interact with and chat about your data via Telegram, WhatsApp, and other messaging platforms
- Control where the OpenClaw agent runs and which models and services it connects to
How Does Adzviser Work with OpenClaw?
Adzviser acts as an MCP server that gives OpenClaw secure, on-demand access to your marketing data. Here's how it works:
- You prompt OpenClaw to connect to the Adzviser MCP server at
https://mcp.adzviser.com/http(see how) - OpenClaw authenticates via OAuth — you log in to Adzviser through your browser to authorize the connection
- You query your data conversationally — ask questions about your Google Ads, Meta Ads, Shopify, and other supported platforms directly in your terminal
Where your data is processed
- Your computer or server runs the OpenClaw agent and coordinates its model requests and tool calls.
- Adzviser and your connected platforms handle data retrieval. Requests go to Adzviser's hosted MCP service at
https://mcp.adzviser.com/http; Adzviser retrieves the requested data from platforms such as Google Ads or Shopify and returns the results to OpenClaw. - Your model provider processes the prompts and tool results included in the model's context. If you use a cloud model, that content is sent to the provider. See how OpenClaw builds model context.
- Messaging services, when enabled, carry the messages and reports you send through them. Using Telegram or WhatsApp adds that service to the data flow.
Self-hosting gives you control over the OpenClaw agent's environment. You can also configure a local model to run model processing on your own infrastructure, but Adzviser data requests still use its hosted MCP service and the connected platforms. Self-hosting does not mean all data stays on your machine.
Why Use OpenClaw with Adzviser?
| Feature | ChatGPT/Claude | OpenClaw |
|---|---|---|
| Runs in terminal | ❌ | ✅ |
| Open-source | ❌ | ✅ |
| Chat via Telegram/WhatsApp/iMessage | ❌ | ✅ |
| Self-hosted agent runtime | ❌ | ✅ |
Example Use Cases
- Daily Ad Spend Check: Query yesterday's spend across Google Ads and Meta Ads right from your terminal
- Telegram Reports: Forward AI-generated marketing summaries directly to a Telegram bot for your team
- Quick Data Lookups: Get campaign-level metrics without opening a browser or dashboard
- Multi-Platform Analysis: Compare performance across platforms with a single prompt
Ready to get started? Follow our MCP Integration Guide to connect Adzviser to OpenClaw.