2.5k
GitHub Stars
2
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill openclaw/skills --skill amplitude-automation- _meta.json298 B
- SKILL.md7.8 KB
Overview
This skill automates common Amplitude product-analytics tasks through the Rube MCP (Composio) toolkit. It handles sending events, resolving users, retrieving user activity, managing cohorts, and browsing event categories while ensuring you always fetch current tool schemas first. Designed for workflow automation where Amplitude operations must be reliable and asynchronous behaviors are handled explicitly.
How this skill works
The skill delegates actions to Rube MCP tool endpoints: it searches available tools first to get up-to-date schemas, verifies an active Amplitude connection, and then sequences toolkit calls for each workflow. Core operations include AMPLITUDE_SEND_EVENTS, AMPLITUDE_FIND_USER, AMPLITUDE_GET_USER_ACTIVITY, AMPLITUDE_IDENTIFY, cohort endpoints, and event-category listing. Many operations are asynchronous (cohort updates, bulk ingestion), so the skill returns request IDs and status-check patterns to confirm completion.
When to use it
- Send tracked events from your app or backend to Amplitude in batches.
- Resolve your application user_id to Amplitude’s internal user ID before fetching activity.
- Update or set user properties using identify operations with $set, $add, $append, etc.
- Add or remove users from behavioral cohorts and track async progress.
- Discover available event categories to validate event_type values before ingestion.
Best practices
- Always call RUBE_SEARCH_TOOLS first to load current tool schemas and connection state.
- Ensure at least one of user_id or device_id is present when sending events or identifying users.
- Use millisecond (13-digit) timestamps for event time fields to avoid date errors.
- Batch events up to the tool schema limit to reduce rate-limit risk and improve throughput.
- For cohort updates, capture request_id and poll CHECK_COHORT_STATUS until complete or error.
Example use cases
- Ingest a batch of page_view and purchase events from a mobile SDK to Amplitude.
- Lookup a customer by email, extract Amplitude internal ID, and fetch their recent event stream.
- Set onboarding properties ($setOnce) and increment a login counter ($add) with IDENTIFY.
- Move a list of churned users into a re-engagement cohort and verify the async update.
- Query event categories to validate event_type casing before sending analytics events.
FAQ
Call RUBE_SEARCH_TOOLS to retrieve current tool schemas and verify the Amplitude connection is ACTIVE via RUBE_MANAGE_CONNECTIONS.
Why are cohort updates asynchronous?
Cohort membership changes can be processed offline for large sets; the API returns a request_id you must poll with CHECK_COHORT_STATUS to confirm completion.