Best Optimizely Experimentation MCP Server MCP Server Alternatives 2026

ClickHouse MCP Server is ClickHouse's official MCP server (ClickHouse/mcp-clickhouse) that connects Claude, Cursor, and other MCP clients to a ClickHouse cluster for fast analytical querying over natural language. Its primary tool, run_query, executes arbitrary SQL against your cluster in read-only mode by default (CLICKHOUSE_ALLOW_WRITE_ACCESS=false) so an AI assistant can explore tables, aggregate billions of rows, and answer analytics questions without risk of mutating data — writes can be enabled explicitly when needed. Companion tools list databases and tables and return schema metadata (including the full create_table_query, with an option to omit per-column detail for lighter responses). A second tool set embeds chDB, ClickHouse's in-process engine, via run_chdb_select_query, letting the assistant query files, URLs, and external databases directly without an ETL step (enabled with the optional mcp-clickhouse[chdb] extra). Destructive statements are gated a second time: even with writes enabled, DROP and TRUNCATE require CLICKHOUSE_ALLOW_DROP=true as well. The server supports both stdio and HTTP/SSE transports; on HTTP/SSE authentication is required rather than optional — startup fails unless a static bearer token (CLICKHOUSE_MCP_AUTH_TOKEN), a FastMCP OAuth/OIDC provider (Azure Entra, Google, GitHub, WorkOS via FASTMCP_SERVER_AUTH), or an explicit local-development opt-out (CLICKHOUSE_MCP_AUTH_DISABLED) is configured. Connection is configured through CLICKHOUSE_HOST, CLICKHOUSE_PORT, CLICKHOUSE_USER, and CLICKHOUSE_PASSWORD, with ClickHouse Cloud, self-hosted, and the public SQL playground all supported.

The MCP server is free and open-source. ClickHouse Cloud: Free trial available. Pay-as-you-go pricing. Self-hosted ClickHouse is free and open-source.

The Sentry MCP Server is Sentry's official Model Context Protocol integration, purpose-built for human-in-the-loop coding agents like Claude Code, Cursor, and Windsurf. Rather than exposing every Sentry API endpoint, it focuses tightly on developer debugging workflows: searching and triaging issues, pulling stack traces and event details, inspecting performance traces, and querying project/team/org metadata in natural language. The primary deployment is a hosted remote MCP server at mcp.sentry.dev, built on Cloudflare's remote-MCP infrastructure, so most users connect with zero local setup — just add the remote URL to their client. For self-hosted Sentry instances or local development, a stdio transport is also available via npx @sentry/mcp-server, authenticated with a Sentry User Auth Token scoped to org:read, project:read, project:write, team:read, team:write, and event:write. AI-powered search tools (search_events, search_issues) translate natural-language queries into Sentry's query syntax, but require a configured LLM provider (OpenAI, Azure OpenAI, Anthropic, or OpenRouter) — all other tools work without one. Claude Code users can also install it as a plugin (claude plugin install sentry-mcp@sentry-mcp) for automatic subagent delegation whenever a conversation touches Sentry errors, issues, or traces. This turns "why did this deploy break in production" into a direct conversational debugging session instead of tab-switching into the Sentry dashboard.

The MCP server is free and open-source. Sentry: Developer tier (free, 5K errors/mo). Team: $26/mo. Business: $80/mo. Enterprise: Custom.

The Datadog MCP Server is Datadog's official, vendor-hosted Model Context Protocol endpoint — not a package. Each Datadog site has its own URL of the form https://mcp.<your-site>/api/unstable/mcp-server/mcp (US1: mcp.datadoghq.com, EU1: mcp.datadoghq.eu), and OAuth 2.0 is the recommended way in; a Personal or Service Access Token as an Authorization bearer header is the documented fallback for CI, with DD_API_KEY plus DD_APPLICATION_KEY headers as a third option. Datadog ships first-party client integrations rather than expecting hand-written config: a Claude Code plugin (/plugin install datadog@claude-plugins-official, then /ddsetup and /ddtoolsets), a Claude connector from the Connectors Directory, plugins for Cursor, VS Code/Copilot, JetBrains and OpenCode, and a ChatGPT app in Preview for US1. Tools are grouped into toolsets selected with a ?toolsets= query parameter, and only `core` — logs, metrics, traces, dashboards, monitors, incidents, hosts, services, events, notebooks — loads by default; two dozen more cover alerting, DBM, DDSQL, RUM, profiling, security, Kubernetes, error tracking, feature flags, cost management and data observability, with apm, cases, code-exec and remote-actions in Preview and excluded from toolsets=all. Access requires the mcp_read or mcp_write role permission in addition to the normal resource permission, which is why a working connection can still return no data. Limits at time of writing are 50 requests per 10 seconds of tool-call burst and 50,000 tool calls per month, and the server is not GovCloud compatible.

The MCP server is free and open-source. Datadog: Free tier (5 hosts). Pro: $15/host/mo. Enterprise: $23/host/mo. Additional products priced separately.

The official Grafana MCP server connects Claude and other AI assistants directly to your Grafana instance and its surrounding observability ecosystem, turning natural-language questions into dashboard lookups, incident investigations, and datasource queries. Dashboard tools cover search, retrieval, JSONPath-scoped property extraction, patch-based editing, and per-panel query/datasource introspection, with context-window-aware helpers like get_dashboard_summary so an agent never has to pull a full multi-megabyte dashboard JSON just to answer a simple question. Query tools speak PromQL against Prometheus (including histogram-percentile helpers), LogQL against Loki, and native query languages for InfluxDB, ClickHouse, CloudWatch, Graphite, Athena, Snowflake, Elasticsearch/OpenSearch, and Quickwit datasources — most gated behind opt-in --enabled-tools flags to keep the default tool surface lean. It also wraps Grafana Incident for creating and updating incidents, Sift for automated error-pattern and slow-request investigations, full alerting CRUD (rules, contact points, notification policies) across Grafana-managed and external Alertmanager sources, Grafana OnCall schedule/shift/alert-group management, RBAC-gated admin tools for teams/users/roles, deeplink generation so the LLM never has to guess a dashboard URL, annotations, snapshots, PNG rendering via the Grafana Image Renderer, and provisioning-repo validation for git-sync workflows. Authentication is a Grafana service account token (Editor role, or granular RBAC scopes) passed as GRAFANA_SERVICE_ACCOUNT_TOKEN alongside GRAFANA_URL, and every tool category can be individually disabled to control context-window usage. On install, the recommended route is uvx: `uvx mcp-grafana` pulls the PyPI package mcp-grafana, which is published by Grafana Labs from this same repository — so despite the server being written in Go, the copy-paste command most Claude Desktop and Cursor configs use is a Python-tooling one, not a binary download. The alternatives are `go install github.com/grafana/mcp-grafana/cmd/mcp-grafana@latest` for a real binary, or the grafana/mcp-grafana container — `-t stdio` for local clients, or the default HTTP mode on :8000 (add `-t streamable-http`) with MCP_GRAFANA_SERVER_TOKEN set to authenticate callers when you expose it.

The MCP server is free and open-source. Grafana Cloud: Free tier (10K metrics, 50GB logs). Pro: $29/mo. Advanced: $299/mo. Self-hosted is free.

Query and analyze your Axiom logs, traces, and all other event data in natural language.

The MCP server is free and open-source. Axiom: Free tier (500GB ingest/mo). Team: $25/mo. Enterprise: Custom pricing.

There is no official SendGrid MCP server, and that is the first thing worth knowing if you searched for one. Twilio owns SendGrid and does publish an official MCP server at twilio-labs/mcp, but that server is generated from Twilio's own Public API specs and covers none of SendGrid's v3 email endpoints — connecting it will not give an assistant a way to send an email or touch a contact list. Twilio's own developer blog acknowledges the gap by publishing a walkthrough on building a SendGrid MCP server yourself rather than shipping one. This entry therefore tracks the most established community implementation, Garoth/sendgrid-mcp (TypeScript, 29 stars), which wraps the SendGrid v3 Marketing API. Its tool surface is broader than most email servers: contact management (list_contacts, add_contact, delete_contacts, get_contacts_by_list), list management (list_contact_lists, create_contact_list, delete_list, add_contacts_to_list, remove_contacts_from_list), sending (send_email for one-off transactional mail with optional dynamic-template data, and send_to_list which creates a SendGrid Single Send against one or more list IDs), dynamic templates (create_template, list_templates, get_template, delete_template), plus get_stats, validate_email, list_verified_senders and list_suppression_groups. Those last two exist because SendGrid rejects sends from unverified senders and requires either a suppression_group_id or a custom_unsubscribe_url on a Single Send, so an agent has to look both up before it can complete a campaign. The README is explicit that only v3 APIs are supported — legacy templates will not work. Install is from source: git clone, npm install, then point your client at the built entrypoint with SENDGRID_API_KEY in the environment. Two cautions. The npm package named sendgrid-mcp is a different project (deyikong/sendgrid-mcp, 3 stars, which does offer a READ_ONLY mode) — do not assume npx installs the repo described here. And this server exposes delete_contacts, delete_list and delete_template, which are irreversible against a live marketing database; scope the API key narrowly rather than granting full access.

The MCP server is free and open-source. SendGrid (by Twilio): Free tier (100 emails/day). Essentials: $19.95/mo. Pro: $89.95/mo. Premier: Custom.

Real User Monitoring data from Datadog.

The MCP server is free and open-source. Datadog RUM: From $1.50/1K sessions/mo. Session Replay: additional $1.80/1K sessions. 14-day free trial.

Mixpanel MCP server for read-only product analytics. Mixpanel itself publishes no MCP server — its agent tooling ships as the mixpanel-headless Python CLI and a set of Claude Agent Skills — so the de-facto server for this API is dragonkhoi/mixpanel-mcp, the community project with the most adoption at 19 stars and 19 forks. It exposes 19 tools over Mixpanel's Query API and nothing over the ingestion API, so an assistant can read your data but cannot write events or edit profiles. Coverage is broad for a single-file server: saved Insights, Funnel and Retention reports (query_insights_report, query_funnel_report, query_retention_report, plus list_saved_funnels and list_saved_cohorts to find the IDs), five segmentation tools including numeric bucketing and per-unit-time sum and average, event and property discovery (get_top_events, get_today_top_events, top_event_properties, top_event_property_values), profile lookup via query_profiles and profile_event_activity, and custom_jql for anything the typed tools do not cover. Authentication is a Mixpanel Service Account username and password plus a project ID, sent as HTTP Basic on every request. There is no published npm package: the project's package.json claims the name mcp-mixpanel, but that name was unpublished from npm in April 2025 and has not existed since, so install is either a git clone and build or the Smithery bundle. Every API URL is hardcoded to mixpanel.com, which means EU- and India-residency projects are not supported. Last commit March 2025.

The MCP server is free and open-source. Mixpanel: Free tier (20M events/mo). Growth: from $28/mo. Enterprise: Custom pricing.

Search, analyze, and query charts, dashboards, experiments, and feature flags.

The MCP server is free and open-source. Amplitude: Starter (free). Plus: from $49/mo. Growth: Custom pricing. Enterprise: Custom.

The PostHog MCP Server is PostHog's official Model Context Protocol integration, giving AI assistants direct access to product analytics, feature flags, session replay, experiments, and error tracking without leaving the chat. It's hosted remotely at mcp.posthog.com (Streamable HTTP) and authenticated with a personal PostHog API key passed as a Bearer token — the quickest setup is `npx @posthog/wizard@latest mcp add`, which auto-configures Cursor, Claude, Claude Code, VS Code, or Zed in one command; manual setup adds an `mcp-remote` proxy entry with the `Authorization` header for clients without native remote-MCP support. Tools cover the full PostHog surface: creating and toggling feature flags with percentage rollouts and targeting rules, running trends/funnel/retention queries via `query-run`, inspecting session recordings, pulling error-tracking issues, and managing experiments — all scoped to the project tied to your API key. Typical use: ask Claude to "create a feature flag for the new checkout flow at 20% rollout" or "how many unique users signed up in the last 7 days, broken down by day?" and the assistant executes the query or mutation against your live PostHog project and returns formatted results. Originally shipped as the standalone PostHog/mcp repo (150 stars), that repository is now archived on GitHub — last pushed 2026-01-19 — with the source moved into the main PostHog monorepo under `services/mcp`. The archive is a relocation, not a retirement: the hosted server and the wizard install are both current, so the link here is a pointer to history and the docs at posthog.com/docs/model-context-protocol are the live reference.

The MCP server is free and open-source. PostHog: Free tier (1M events/mo). Pay-as-you-go after. Self-hosted is free and open-source.