Indexes the coding-agent sessions already sitting on your disk — months of history from before you installed it — and serves them back over MCP. 3.5 GB searched in ~1.5 ms, no LLM, no embeddings.
Developer ToolsDatabases & Data
1 views
amplitudeproduct-analyticsexperimentationfeature-flagscohortssession-replayoauth
Summary
Amplitude's official hosted MCP server with 50+ tools over analytics, cohorts, experiments, feature flags, taxonomy and session replay, secured by OAuth and existing RBAC roles.
Features
- 50+ tools spanning analytics, content, cohorts, experiments, taxonomy and session replay
- Ad-hoc segmentation, funnel and retention queries, plus chart creation and rendering
- Build dashboards and notebooks, and manage comments and sharing
- Create and sync cohorts, read user profiles and activity timelines
- Design and query A/B tests, define reusable KPIs, manage feature flags and deployments
- Read/write taxonomy: events, properties, transformations, group types and tracking-plan branches
- OAuth 2.0 with RBAC — separate read and write MCP actions; US and EU residency endpoints
Installation
Set up this MCP server in your favorite AI agent — copy a ready-made configuration below.
Description
Amplitude's official MCP server is a hosted endpoint — https://mcp.amplitude.com/mcp for US accounts, https://mcp.eu.amplitude.com/mcp for EU data residency — that exposes more than 50 tools across the product analytics surface. Setup is OAuth 2.0 and takes about two minutes; there are no API keys to manage.
The breadth is what distinguishes it from a query-only analytics connector:
- Analytics — ad-hoc segmentation, funnel and retention queries; create and render charts; manage chart monitors and alerts.
- Content — build dashboards and interactive notebooks, manage comments and sharing.
- Cohorts and users — create and sync cohorts, read user profiles and activity timelines, export cohort data.
- Experiments and flags — design and query A/B tests, define reusable metrics, manage feature flags and deployments.
- Taxonomy — read and write events, properties and transformations, manage group types and tracking-plan branches.
- Plus session replay retrieval and analysis, guides and surveys, customer feedback analysis, AI agent performance analytics, and data-warehouse connection and job history.
Access is governed by Amplitude's existing RBAC rather than a separate permission model: Use MCP (read) covers querying and retrieval, Use MCP (write) covers creating and editing objects and tracking-plan changes, and Member, Manager and Admin roles already include write. Data visibility is limited to projects your account can already see. All organisation users have access by default unless an admin restricts it.
Appending ?discovery=progressive to either endpoint enables progressive tool discovery, which keeps the tool list from flooding a client's context.
Two limits to note: it is not an ingestion endpoint — production events still go through the SDKs or the HTTP V2 API — and public documentation lives behind a separate Docs MCP server. Third-party AI services process whatever data you send, under their own terms.
Best for product managers and growth teams who want to interrogate funnels, ship a flag and update the tracking plan in one conversation.
Related MCP Servers
Weaviate's MCP server, built into the database itself — hybrid search, schema inspection, tenant listing and batch upserts at /v1/mcp, scoped by RBAC.
Databases & Data
Google's open-source MCP server for enterprise databases — one binary that fronts AlloyDB, BigQuery, Spanner, Postgres, MySQL, Oracle, MongoDB, Snowflake and more.
Databases & DataCloud & Infrastructure
4 views
A universal database MCP server — one integration for PostgreSQL, MySQL, SQL Server, SQLite, and MariaDB.
Databases & Data