MCP App Store
by KeenEthics
Flowsery - web analytics and session insights icon

Flowsery - web analytics and session insights MCP Server for Claude and ChatGPT

By Flowsery

Overview

Flowsery is cookie-free web analytics. It tracks traffic, goals and revenue for your websites, and its AI reviews session recordings to build a ranked list of issues: bugs, broken flows and UX problems, each with how many sessions hit it and steps to replicate. From Claude you can: list the websites in your workspace; get an overview of visitors, pageviews, bounce rate and revenue for any date range, or as a timeseries; break traffic down by page, referrer, channel, campaign, country, region, city, device, browser, operating system or hostname, or combine dimensions in a custom breakdown; see who is on the site right now and where they are; follow a single visitor's journey step by step; pull the top AI-detected issues, open one with its occurrences and affected sessions, and mark it resolved once fixed; and record or delete goal events and payments so revenue is tied back to the channel that brought it. Twenty-seven tools in total. Sign in with your Flowsery account. You need at least one website with the Flowsery tracking script installed.

How to connect

  1. Claude: Open the Flowsery - web analytics and session insights listing in the Claude connectors directory and choose Connect. Claude listing Flowsery - web analytics and session insights exposes 27 tools here.
  2. ChatGPT: Open the Flowsery - web analytics and session insights listing in the ChatGPT app directory and choose Connect. ChatGPT listing Flowsery - web analytics and session insights exposes 28 tools here.

Tools

delete_goals

ChatGPT
Permanently delete recorded goal completions matching every filter given (filters combine with AND). At least one of visitorId, name, startAt, or endAt is required or the call fails before reaching the API; startAt and endAt are independent, so one bound alone is allowed. Without a date range, matches are deleted across the whole history. Returns the number of rows deleted. Cannot be undone: restate website, filters, and range and get explicit confirmation first. Deletes completions only; the goal definition stays and get_goals still lists it. Use update_issue_status for issues. Requires websiteId or domain with a workspace token.

delete_payments

ChatGPT
Permanently delete analytics payment records, without issuing refunds or changing a payment provider. Deletes records matching every filter given (filters combine with AND): one transactionId, all payments of a visitorId, and/or a creation window. At least one of transactionId, visitorId, startAt, or endAt is required or the call fails before reaching the API; startAt and endAt are independent, so one bound alone is allowed. Without a date range, matches are deleted across the whole history. Returns the number of rows deleted. Cannot be undone and removes the revenue from every report and visitor profile, so restate website, filters, and range and get explicit confirmation first. To record a refund that already happened, use track_payment with isRefund instead. Requires websiteId or domain with a workspace token.

get_breakdown

ChatGPT
Group visitors by any one of 25 dimensions, ranked by visitors descending, for a date range. Generic form of the named get_ breakdown tools: use it for dimensions without one (entry_page, exit_link, browser_version, os_version, utm_source, utm_medium, utm_term, utm_content, ref, source, via, all_params); for page, referrer, country, region, city, device, browser, os, campaign, hostname, channel, or goal the dedicated tool returns the same rows. Combine dimension with filter_ to drill in: dimension page plus filter_utm_campaign shows where one campaign landed. Rows carry value, visitors, revenue, and percentage with pagination.total; limit defaults to 100 (max 1000). Dates default to the last 30 days; all filter_* arguments apply. Requires websiteId or domain with a workspace token.

get_browsers

ChatGPT
Get visitors grouped by browser name (Chrome, Safari, Firefox, Edge, and others), ranked by visitors descending, for a date range. Names only: use get_breakdown with dimension browser_version for versions, get_operating_systems for the OS split, and get_devices for desktop versus mobile. Pass filter_browser to other tools to scope them to one browser. Rows carry value, visitors, revenue, and percentage with pagination.total; limit defaults to 100 (max 1000). Dates default to the last 30 days; all filter_* arguments apply. Requires websiteId or domain with a workspace token.

get_campaigns

ChatGPT
Get UTM campaigns (utm_campaign values) ranked by visitors descending for a date range. Only visits tagged with utm_campaign appear, so untagged traffic is absent; use get_referrers or get_channels for the full source picture. Use get_breakdown with dimension utm_source, utm_medium, utm_term, utm_content, or all_params for the other tracking parameters. Rows carry value, visitors, revenue, and percentage with pagination.total; limit defaults to 100 (max 1000). Dates default to the last 30 days; all filter_* arguments apply. Requires websiteId or domain with a workspace token.

get_channels

ChatGPT
Get visitors grouped into GA4-aligned marketing channels (Organic Search, Paid Search, Organic Social, Paid Social, Email, Display, Referral, Direct, Affiliate, Video, SMS, Audio), ranked by visitors descending, for a date range. Channels are classified from the referrer domain and utm_medium or utm_source. Use this first for the traffic mix, then get_referrers for the domains behind Referral and Organic Social, or get_campaigns for tagged campaigns. Rows carry value, visitors, revenue, and percentage with pagination.total; limit defaults to 100 (max 1000). Dates default to the last 30 days; all filter_* arguments apply. Requires websiteId or domain with a workspace token.

get_cities

ChatGPT
Get visitors grouped by city, ranked by visitors descending, for a date range. Finest geographic tool and the longest tail: pass filter_country or filter_region first so the top rows are meaningful, and raise limit above the default 100 (max 1000) when you need more. Use get_countries or get_regions for a coarser view. Rows carry value, visitors, revenue, and percentage with pagination.total. Dates default to the last 30 days; all filter_* arguments apply. Requires websiteId or domain with a workspace token.

get_countries

ChatGPT
Get visitors grouped by country, ranked by visitors descending, for a date range. Coarsest of the three geographic tools: use get_regions for states or provinces and get_cities for cities, and add filter_country to either to drill into one country. Use get_realtime_map for where visitors are right now instead of over a range. Rows carry value, visitors, revenue, and percentage with pagination.total; limit defaults to 100 (max 1000). Dates default to the last 30 days; all filter_* arguments apply. Requires websiteId or domain with a workspace token.

get_devices

ChatGPT
Get visitors split by device type (Desktop, Mobile, Tablet), ranked by visitors descending, for a date range. Use this for the mobile-versus-desktop question; use get_browsers or get_operating_systems for the software split. Pass filter_device to any other tool to restrict it to one device type instead. Three rows at most, so pagination rarely matters. Rows carry value, visitors, revenue, and percentage. Dates default to the last 30 days; all filter_* arguments apply. Requires websiteId or domain with a workspace token.

get_goals

ChatGPT
Get every configured goal (custom events plus the auto-created payment and free_trial goals) with how many visitors completed it in the date range. Use this to compare conversions across goals; use get_overview for conversion_rate against the site's KPI goal, get_breakdown with dimension goal for the same list with revenue and percentage, and get_visitor for one person's completions. Dates default to the last 30 days; filter_* narrows the visitors counted and limit/offset page the goal list. Goals are created by track_goal. Requires websiteId or domain with a workspace token.

get_hostnames

ChatGPT
Get visitors grouped by hostname, ranked by visitors descending, for a date range. Useful when one website tracks several domains or subdomains (www, app, docs); a single-domain site returns one row. Use get_pages for paths within a host, and pass filter_hostname to any other tool to scope it to one host. Rows carry value, visitors, revenue, and percentage with pagination.total; limit defaults to 100 (max 1000). Dates default to the last 30 days; all filter_* arguments apply. Requires websiteId or domain with a workspace token.

get_issue

ChatGPT
Get full detail for one AI-detected issue: every occurrence with timestamps, the sessions behind it, steps to replicate, comments, and any linked Linear or Jira ticket. Get issueId from list_issues; use this only when you need the evidence behind one issue. An unknown id, or one from another website, fails with 'Issue not found'; on a free trial only the first 10 issues are available, and later ones fail as not included in the current plan. Each session lists its recording id, the anonymous visitorUid, start URL and page title, timing and event counts; no names, emails or locations are returned, so surface only what answers the question. Requires websiteId or domain with a workspace token.

get_metadata

ChatGPT
Get one website's settings: domain, timezone, currency, KPI goal name, logo, and color scheme. Read-only; nothing is changed. Call it after list_websites to learn the timezone and currency before running date-range reports, then pass that timezone to the report tools. With a workspace token pass websiteId or domain; with neither it returns the same website list as list_websites, so prefer list_websites for discovery. A website key needs no selector.

get_operating_systems

ChatGPT
Get visitors grouped by operating system (Mac OS, Windows, iOS, Android, Linux, and others), ranked by visitors descending, for a date range. Names only: use get_breakdown with dimension os_version for versions, get_browsers for the browser split, and get_devices for desktop versus mobile. Pass filter_os to other tools to scope them to one OS. Rows carry value, visitors, revenue, and percentage with pagination.total; limit defaults to 100 (max 1000). Dates default to the last 30 days; all filter_* arguments apply. Requires websiteId or domain with a workspace token.

get_overview

ChatGPT
Get headline totals for one website over a date range as a single row: visitors, sessions, bounce rate, average session duration, revenue, revenue per visitor, and conversion rate. Dates default to the last 30 days ending now; timezone defaults to the site setting. Every filter_ argument narrows the whole result, so filter_country plus filter_device answers 'mobile visitors from Germany' in one call. Use get_timeseries for the trend over time and a get_ breakdown tool for the split by page, source, or geography. Requires websiteId or domain with a workspace token.

get_pages

ChatGPT
Get page paths ranked by visitors, descending, for a date range: which pages get the most traffic. Use get_breakdown with dimension entry_page for landing pages or exit_link for outbound clicks, and get_hostnames when the site serves several domains. Add filter_utm_campaign or filter_referrer to see where one source's traffic landed. Rows carry value, visitors, revenue, and percentage with pagination.total; limit defaults to 100 (max 1000). Dates default to the last 30 days; all filter_* arguments apply. Requires websiteId or domain with a workspace token.

get_realtime

ChatGPT
Count the visitors active on the site within the last 5 minutes. A point-in-time number with no history: it takes no date, filter, or pagination arguments. Use get_timeseries with interval hour for recent trends and get_realtime_map when you need where those visitors are. Returns data[0].visitors. Poll no more than once every 5 seconds. Requires websiteId or domain with a workspace token.

get_realtime_map

ChatGPT
Get the visitors active on the user's own website in the last 5 minutes with their approximate city-level location, for a live map view. Coordinates are rounded to 0.1 degrees, and no IP addresses or other identifying details are returned. Use get_realtime when only the count matters, and get_countries or get_cities for geography over a historical date range. Takes only the website selector: no dates, filters, or pagination. Poll no more than once every 5 seconds. Requires websiteId or domain with a workspace token.

get_referrers

ChatGPT
Get referrers ranked by visitors, descending: which external sites sent traffic in the date range. Recognized sites appear under a source name (Google, ChatGPT, Reddit), others under their domain. Use get_channels when you want traffic grouped into GA4-style channels (Direct, Organic Search, Paid Social) instead of individual domains, and get_campaigns or get_breakdown with dimension utm_source for traffic identified by UTM tags rather than referrer. Rows carry value, visitors, revenue, and percentage with pagination.total; limit defaults to 100 (max 1000). Dates default to the last 30 days; all filter_* arguments apply. Requires websiteId or domain with a workspace token.

get_regions

ChatGPT
Get visitors grouped by region or state name (such as California or Bavaria), ranked by visitors descending, for a date range. Sits between get_countries (coarser) and get_cities (finer); combine with filter_country to list the regions of one country. Pass filter_region to other tools to scope them to one region. Rows carry value, visitors, revenue, and percentage with pagination.total; limit defaults to 100 (max 1000). Dates default to the last 30 days; all filter_* arguments apply. Requires websiteId or domain with a workspace token.

get_timeseries

ChatGPT
Get the same metrics as get_overview bucketed by hour, day, week, or month, plus totals across the whole window. Returns one point per bucket with timestamp, name, visitors, sessions, revenue split into newRevenue, renewalRevenue and refundedRevenue, conversionRate, and kpiValue. Use this for trends and charts; use get_overview for one total and a get_* breakdown tool for a split by dimension rather than time. Dates default to the last 30 days and interval to day. Match interval to range: hourly buckets across a year return thousands of points. Requires websiteId or domain with a workspace token.

get_visitor

ChatGPT
Get the analytics for one visitor to the user's own website: country and city, device and browser, acquisition source, activity (visit and pageview counts, visited pages, completed goals), revenue totals, whether the visitor is identified, and a timeline of pageviews, goals, and purchase amounts, newest first. Names, emails, user IDs, IP addresses, and card or billing details are never returned. Call it only when the user asks about a specific visitor, and share only what answers the question. profile is null for anonymous visitors; each list is capped at the 100 most recent items. Use the aggregate get_* tools for questions about many visitors. visitorId is the visitor record ID from get_realtime_map or the dashboard visitor view, not the _fs_vid cookie value; an unknown id, or one from another website, fails with 'Visitor not found'. Requires websiteId or domain with a workspace token.

list_issues

ChatGPT
List issues the AI found while analyzing session recordings: bugs, broken flows, and UX problems, deduplicated across sessions and ranked by severity (or by last seen with sort recency). Each row has title, severity, status, sessions affected, and first/last seen; the response also carries site-wide open, in_progress, and resolved counts plus pagination.total. Start here for 'what is broken', then call get_issue with an id for occurrences, steps to replicate, and comments. Suspended issues are hidden unless status is suspended, so an issue that vanished was probably suspended, not deleted. On a free trial only the first 10 issues are listed and counted. Limit defaults to 100 (max 1000). Requires websiteId or domain with a workspace token.

list_websites

ChatGPT
List the websites the current API token can read, with id, domain, timezone, currency, and KPI goal per site. A workspace token (flow_ws_) returns every website in the workspace; a website key (flow_) returns only its own. Call this first with a workspace token: every other tool then needs websiteId or domain from this list, and omitting both fails with 'Website ID or domain is required'. Takes no parameters. Use get_metadata instead when you already know the website and only need its settings.

list_workspaces

ChatGPT
List the workspaces this sign-in can act in, across every organization the user belongs to. Returns { status, data } with one entry per workspace: id, name, organization { id, name }, role { key, name }, permissions, isDefault and current. Call this when the user names a workspace, client or organization, or when data they expect is missing, then pass the matching id as workspaceId to every other tool. Without workspaceId, tools act in the workspace marked current. A workspace API token (flow_ws_) belongs to one workspace, so it lists only that one; a website key (flow_) belongs to one website and fails with a 400 here. Takes no arguments.

track_goal

ChatGPT
Record one completion of a custom goal. The goal is created on first use, so no setup call is needed; names are lowercase letters, digits, underscores, and hyphens, max 64 chars. Pass visitorUid (the anonymous visitor ID Flowsery assigned on the user's own website) so the completion attaches to that visitor's sessions and source; omit it for an anonymous completion. Each call appends a completion, so repeating it counts the goal twice. Undo with delete_goals. Requires websiteId or domain with a workspace token.

track_payment

ChatGPT
Record an analytics payment; this never charges a customer, moves money, or issues a refund. Revenue appears in get_overview, get_timeseries, every breakdown's revenue column, and the visitor profile. Skip it when the site's provider (Stripe, LemonSqueezy, Polar, and other connected providers) is tracked automatically, or the revenue is counted twice; use track_goal for conversions without revenue. transactionId must be unique: a repeated id is rejected, not deduplicated. A new payment also records a payment goal completion (free_trial when amount is 0); isRenewal skips that. isRefund with an existing transactionId overwrites that payment's recorded refund amount and any supplied customer attribution instead of adding a row. Attribution looks up a known visitor by visitorUid, customerId, or email; with no match the revenue is kept but its source shows as Unknown. Send email and name only when the user supplied them and wants the payment attributed. Undo with delete_payments. Requires websiteId or domain with a workspace token.

update_issue_status

ChatGPT
Set an issue's status to open, in_progress, resolved, or suspended and return the updated issue. Only the status changes; title, severity, occurrences, and comments stay, and any status can be set again later, so this is reversible. Confirm which state the user means before calling: resolved asserts the bug is fixed, suspended hides a known non-problem from the default list_issues result. Not a delete: issues cannot be removed through this server. Get issueId from list_issues; an unknown id fails with 'Issue not found', and on a free trial issues beyond the first 10 fail as not included in the current plan. Requires websiteId or domain with a workspace token.

delete_goals

Claude

delete_payments

Claude

get_breakdown

Claude

get_browsers

Claude

get_campaigns

Claude

get_channels

Claude

get_cities

Claude

get_countries

Claude

get_devices

Claude

get_goals

Claude

get_hostnames

Claude

get_issue

Claude

get_metadata

Claude

get_operating_systems

Claude

get_overview

Claude

get_pages

Claude

get_realtime

Claude

get_realtime_map

Claude

get_referrers

Claude

get_regions

Claude

get_timeseries

Claude

get_visitor

Claude

list_issues

Claude

list_websites

Claude

track_goal

Claude

track_payment

Claude

update_issue_status

Claude

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/flowsery"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/flowsery)