Automate Amplitude Data Management

Automates Amplitude product analytics through Composio's Amplitude toolkit, connected via Rube MCP. It covers five core workflows:

1. Send Events - AMPLITUDE_SEND_EVENTS posts one or more events, each needing an event_type and either a user_id or device_id; timestamps must be 13-digit millisecond epoch, and events are processed asynchronously (a successful response doesn't mean the data is immediately queryable).
2. Get User Activity - resolves your application's user ID to Amplitude's internal ID via AMPLITUDE_FIND_USER, then retrieves the event stream with AMPLITUDE_GET_USER_ACTIVITY (paginated via offset/limit, reverse chronological by default).
3. Find and Identify Users - AMPLITUDE_FIND_USER searches by user_id, device_id, or Amplitude ID; AMPLITUDE_IDENTIFY sets user properties using $set, $setOnce, $add, $append, and $unset operations.
4. Manage Cohorts - list cohorts, fetch cohort detail, and update membership (add/remove arrays of user IDs) via an async flow: submit the update, then poll AMPLITUDE_CHECK_COHORT_STATUS with the returned request_id until it's complete or error. Only behavioral cohorts support API membership updates.
5. Browse Event Categories - AMPLITUDE_GET_EVENT_CATEGORIES lists configured categories, useful for validating event_type values before sending events.

Use it whenever a Claude Code session needs to send analytics events, look up a user's activity, manage cohort membership, or browse event taxonomy in Amplitude - without hand-writing raw API calls.

Setup requires no API keys: add https://rube.app/mcp as an MCP server, confirm RUBE_SEARCH_TOOLS responds, then call RUBE_MANAGE_CONNECTIONS with toolkit amplitude and complete the returned auth link if the connection isn't already ACTIVE. The skill always calls RUBE_SEARCH_TOOLS first to get current tool schemas before acting.

Teams running Amplitude who want an AI assistant to send events, resolve user activity, and manage cohorts directly through Rube MCP's Composio toolkit rather than the Amplitude API or dashboard.

• User IDs: Amplitude keeps its own internal user IDs separate from your application's. AMPLITUDE_FIND_USER resolves your ID to Amplitude's internal ID; AMPLITUDE_GET_USER_ACTIVITY requires that internal ID, not your application's user_id.
• Event timestamps: must be milliseconds since epoch (13 digits) - a 10-digit seconds value is misread as a very old date. Omitting the timestamp uses server receive time.
• Rate limits: event ingestion has per-project throughput limits, so batching events reduces API calls; cohort membership updates have their own async processing limits.
• Response parsing: response data may be nested under a data key; user activity returns events in reverse chronological order; cohort lists may include archived cohorts (check the status field); parse defensively with fallbacks for optional fields.