Mode Analytics MCP. Manage reports and audit SQL queries from chat.

This MCP Server lets your AI agent take full control of Mode Analytics workflows, letting you talk to your data instead of clicking through dashboards. You'll never have to manually navigate workspaces or run queries in a web interface again; your agent handles it all via natural conversation.

get_mode_account: Retrieves the basic details for the authenticated Mode Analytics account, so your agent knows exactly which workspace you're dealing with.

list_mode_spaces: Lists every individual data workspace—or 'space'—available in your account. You can use this tool to inventory all the different areas where your team keeps its analysis and reports.

list_mode_members: Provides a list of every user member belonging to the Mode Analytics workspace, letting you check who has access to what data.

list_mode_reports: Finds and lists all available reports within a specified space. You can point your agent to a specific area and get a catalog of everything that's ready for review or execution.

get_mode_report: Fetches detailed metadata for a single, specific report in your workspace. This includes checking the report's parameters and understanding its current status without having to open it up manually.

list_mode_definitions: Gives you a list of all calculated field definitions used across your reports. You use this tool to verify that metrics are being computed consistently everywhere, which is crucial for making sure everyone's using the same numbers.

When auditing data pipelines and understanding exactly what code drives your reports, list_mode_queries retrieves and lists all underlying SQL queries associated with a given report’s definition. This lets you see the raw logic—the exact SQL—so you know where every number is actually coming from.

For checking how far along a specific run is, get_mode_report_run gets the current status and history of a particular executed report run. If you need to check the overall timeline for a single report ID, list_mode_report_runs lists the entire history of executed reports for that ID, showing the status changes and timestamps.

If you're ready to compute something new or rerun an old analysis, run_mode_report starts a brand-new report run using defined parameters. This tool triggers the necessary data computation on the fly, letting your agent support custom inputs right when it executes.

Your AI client can also use these functions in conjunction with get_mode_report and list_mode_reports. You'll tell your agent which space you want to work in, then ask it to list all reports there. Once the report is identified, the agent pulls its specific metadata so you know what parameters are needed before running it.

If those parameters change, or if someone modifies the underlying SQL logic—which list_mode_queries shows—your agent keeps you informed.

The combination of these tools means your AI client can manage everything from initial discovery and user auditing to complex execution and deep code inspection. You're not just reading data; you're running the whole engine.