# Amazon Marketing Cloud MCP for AI Agents MCP > The Amazon Marketing Cloud MCP connects your AI agent directly to advanced advertising analytics, letting you manage complex data workflows and run ad-hoc SQL queries through natural conversation. You can monitor execution status, create new reporting pipelines, and instantly retrieve secure download links for cross-channel attribution reports without touching the AWS console. ## Overview - **Category:** industry-titans - **Price:** Free - **Tags:** clean-room, sql-queries, cross-channel-analytics, privacy-safe, data-pipelines, marketing-intelligence ## Description Managing large datasets in Amazon Marketing Cloud usually means navigating multiple dashboards or writing detailed SQL scripts just to get a simple answer. This MCP changes that process entirely. Now, your agent handles the entire data pipeline through plain conversation. You can ask it to list all existing reporting workflows, trigger a complex multi-touch attribution query for last month, and then wait for results—all by talking to it. When the query finishes, you don't have to hunt down where the files are; your agent retrieves the secure download URL right in the chat. Plus, if something stalls or takes too long, you can ask it to check the status or even cancel the job entirely. The Vinkius catalog makes this connection simple: connect your preferred AI client and start generating enterprise-grade analytics instantly. ## Tools ### list_workflows Retrieves a list of all saved Amazon Marketing Cloud reporting pipelines. ### create_workflow Sets up and saves a new, reusable data analysis workflow in AMC. ### get_workflow_details Pulls detailed information about a specific existing reporting pipeline. ### delete_execution Stops and cancels a running or pending data job immediately. ### update_workflow Modifies the parameters or structure of an existing reporting pipeline. ### delete_workflow Permanently removes a defined AMC workflow from your account. ### list_executions Shows a history and list of all past data job runs. ### create_execution Starts a new, one-time ad-hoc SQL query or workflow run. ### get_execution_status Checks if a running data job is PENDING, COMPLETED, or failed. ### get_execution_download_urls Retrieves the specific secure URLs needed to download completed query results. ## Prompt Examples **Prompt:** ``` List all my saved workflows. I want to see if I already have one for Q3 path analysis. ``` **Response:** ``` **📚 Available Reporting Workflows** * `Campaign Overlap Analysis`: (Last run: 2026-05-15) * `Path to Conversion`: (Last run: 2026-04-28) * `Q3 Path Analysis`: (Status: DRAFT - needs update) ``` **Prompt:** ``` Run the 'Path to Conversion' workflow for May. And then show me the secure download link when it's done. ``` **Response:** ``` **🚀 Execution Triggered** Successfully started execution for `Path to Conversion`. * **Execution ID:** exec-abc-123 * **Status:** PENDING (Processing data) Please wait a few minutes. I'll monitor it and let you know when the download URL is ready. ``` **Prompt:** ``` I need to check the status of execution exec-45678 and, if it finished, get the results for my team. ``` **Response:** ``` **✅ Status Check Complete** Execution ID `exec-45678` is **COMPLETED**. The query ran successfully. Here are the download details: * **File Type:** CSV (Attribution data) * **Download URL:** [secure_download_link]/exec-45678/results Please let me know if you want to schedule this report monthly. ``` ## Capabilities ### Manage Reporting Workflows List, create, update, or delete saved SQL workflows to standardize your advertising reports. ### Run and Track Queries Execute ad-hoc SQL queries or trigger entire workflow pipelines and monitor their processing status (Pending, Completed). ### Audit Execution History View a record of past data jobs, check the status of specific executions, and cancel long-running tasks. ### Retrieve Data Outputs Get immediate, secure download URLs for completed execution results directly in your chat window. ### Orchestrate Complex Analytics Trigger multi-touch attribution queries and manage the entire data flow without needing to access the AWS console. ## Use Cases ### Investigating Campaign Overlap A media buyer needs to understand how different campaign types overlapped in the last quarter. They ask their agent, which triggers a complex multi-touch attribution query, and receive the necessary insights without manually building the SQL. ### Standardizing Monthly Reporting A data analyst needs to run the same five reports every month. Instead of rewriting the process, they use `list_workflows` to verify existing definitions and trigger a new execution via `create_execution`. ### Debugging Stalled Queries A marketing engineer runs a massive query that seems stuck. They check the status using `get_execution_status`, realize it's stalled, and then use `delete_execution` to cancel and restart the job cleanly. ### Quickly Sharing Insights An advertising strategist needs a report for an executive meeting. They prompt their agent to execute a workflow and immediately get the secure download URL, which they can share in minutes. ## Benefits - Run complex attribution analyses instantly. Instead of writing and debugging multi-touch models in a console, your agent handles the query execution for you. - Gain full oversight on data jobs. You can list executions and check status using `get_execution_status` to know exactly when your report will be ready. - Automate reporting setup with confidence. Use `list_workflows` and `create_workflow` to build standard, reusable pipelines that anyone can trigger with a simple command. - Stop wasting time hunting for files. Once the data is done processing, you just ask for it; your agent uses `get_execution_download_urls` to deliver the secure link immediately. - Maintain clean governance. You can use `delete_workflow` or `update_workflow` to keep your reporting definitions standardized and accurate without touching complicated AWS settings. ## How It Works The bottom line is: your AI client acts as a conversational API layer, transforming complex data management into simple chat commands. 1. Subscribe to this MCP, providing your Amazon Marketing Cloud instance ID, client credentials, region, and profile details. 2. Once connected, tell your agent what data you need. You can ask it to list existing workflows or execute a brand-new ad-hoc query with specific parameters (like date ranges). 3. Your agent manages the job in the background. When the query is finished, you simply ask for the results, and it provides the secure download URL. ## Frequently Asked Questions **How does the Amazon Marketing Cloud MCP help me with cross-channel attribution?** It lets you run complex, multi-touch attribution queries that would otherwise require extensive SQL writing. You simply ask your agent for the analysis, and it handles the entire data execution process. **Can I use this MCP if I don't know the exact SQL syntax?** Yes. This tool translates your natural language requests into structured commands. You tell it what you need—like 'overlap between paid social and search'—and the agent figures out the right query. **What if my data job fails or gets stuck?** You can monitor every job through your AI client. If something stalls, you can use the MCP to check its status or even cancel and restart the execution easily, saving time. **Does this Amazon Marketing Cloud MCP only run reports I've already set up?** No. Besides managing existing workflows, you can also trigger brand new, ad-hoc queries on demand for one-off analysis whenever required. **How do I get the final report from this MCP? Is it a file download?** When the query finishes successfully, the MCP doesn't just tell you it worked; it provides a secure, direct download URL right in your chat, ready for you to use.