MCP Connector

The Hive Analytics MCP (Model Context Protocol) server gives any compatible AI assistant a standard, secure way to work with your Hive analytics data. Connect it to Claude, ChatGPT or Microsoft Copilot, and ask about your events, viewers and networks in plain language.

The server is hosted by Hive and follows the authenticated remote MCP specification, so there is nothing to install. You sign in with your Hive account, and the connector provides read-only tools for event performance, viewer locations and networks, and trends over time.

https://mcp.hivestreaming.com/

For ideas on where to start, see the example prompts in Common use cases.

How connecting worksCopy Anchor Link

Your assistant connects to the server over the web (Streamable HTTP). Setup is the same two steps on every platform:

1. An admin makes the connector available to your organization in Claude, ChatGPT or Microsoft 365. On personal plans, you are the admin.
2. Each person connects their own Hive account by signing in through Hive Auth. Answers are scoped to that account.

Before you beginCopy Anchor Link

• A Hive account with access to IT & Network Analytics (INA)
• A supported plan for your AI assistant
• Your AI workspace admin, if your organization manages connectors centrally

Pick your assistantCopy Anchor Link

For any client, add the server URL https://mcp.hivestreaming.com/ (Streamable HTTP), then sign in with your Hive account when prompted.

Common use casesCopy Anchor Link

Copy an example prompt into your own chat to see what's possible, then replace the placeholders in brackets.

Post-event reportCopy Anchor Link

Turn a finished event into a short report you can share with stakeholders: turnout, reliability and where people watched from.

How did [yesterday's town hall] go?

Using Hive Analytics, pull the total number of viewers for [event name] on [date], how long the event ran, its reliability score, and the top 10 offices or sites by viewer count.

Write a summary of no more than 200 words for [audience, e.g. our internal comms team], followed by a short table of the top sites. Call out anything that looks unusual, such as sites with low reliability or a large share of viewers with an unknown location.

Reliability checkCopy Anchor Link

Find events that didn't go well before someone else asks. Useful for IT and network teams reviewing the last few weeks.

Did any events in the last 30 days have a reliability score below 3?

Using Hive Analytics, list every event in the last [30 days] with its date, viewer count and reliability score.

Sort the events from lowest to highest reliability. For each event below 3, show which sites had the most viewers so I know where to start investigating.

Site and network investigationCopy Anchor Link

Dig into one location after an event: who watched from there, and whether they were on the corporate network or a VPN.

How did viewers in [city] experience [event name]?

Using Hive Analytics, show how many viewers watched [event name] from [site or city], the networks they connected from, and how many were on a VPN.

Compare the site with the event as a whole. Tell me in plain language whether this site looks different from the rest and what I should check next with the local IT team.

Turnout by officeCopy Anchor Link

See how far your messages actually reach. Internal communications teams use this to spot offices that consistently tune in less.

Which offices had the most viewers last quarter?

Using Hive Analytics, look at all events in [Q3 2026] and rank our offices or sites by total viewers.

List the 5 strongest and 5 weakest sites. Then give me 3 short observations I could share with the internal communications team about where our reach is strong and where it falls short.

Summary for leadershipCopy Anchor Link

Summarize a month, quarter or half-year of streaming activity for a business review or leadership update.

How many events did we run in Q3, and how many people watched?

Using Hive Analytics, summarize [Q3 2026] for [our quarterly business review]: number of events, total viewers, average viewers per event, and average reliability. Compare each number with [Q2 2026].

Write 3 to 5 bullet points for an executive audience. Lead with the biggest change, keep each bullet to one line, and avoid technical terms.

Traffic savings by officeCopy Anchor Link

See how much streaming traffic Hive eCDN saved across known network sites. Network teams use this to show the value of eCDN and to spot sites saving less than expected.

How much network traffic did Hive offload at each known network site last month?

Using Hive Analytics, show the traffic saved at each known network site for events in [last quarter], as a percentage and in GB. Rank the sites from highest to lowest savings.

Flag sites with savings below [70%] and include their viewer counts. Finish with a one-paragraph summary I can share with [our network team].

Devices and browsersCopy Anchor Link

See what your audience watches on. Useful for planning browser support and spotting how many people join from mobile.

How many viewers watched on mobile last month?

Using Hive Analytics, break down viewers for events in [last quarter] by device type, then by browser and by operating system.

Show the top 5 in each breakdown with their share of viewers. Call out any browser or operating system with a large share that [our IT team] should make sure is supported.

CDN and peer-to-peer deliveryCopy Anchor Link

Check how video reached viewers: how much went over the CDN, how much was shared peer to peer, and how many CDN requests succeeded.

How much of our event traffic was delivered peer to peer last month?

Using Hive Analytics, show total traffic, CDN traffic and peer-to-peer traffic for events in [last quarter], with the offload share and the CDN success rate.

Compare these figures with [the previous quarter] and summarize the difference in 3 bullet points for [our network team].

Viewing experienceCopy Anchor Link

See how smooth playback was and what video quality people watched in.

How smooth was playback for our events last month?

Using Hive Analytics, group events in [the last 30 days] by playback reliability, from Excellent to Very poor, and show viewing minutes per video resolution.

Tell me what share of viewing was in HD and how many events had poor or very poor reliability.

Good to knowCopy Anchor Link

Reliability is scored from 0 to 5, and anything below 3 is worth a closer look. A location shown as unknown means the network couldn't be identified, not that nobody watched.

By default, answers cover the last 30 days unless you ask for a different period. You can also narrow results to a viewer count range or a specific location when you want to zoom in.

Data, permissions and limitsCopy Anchor Link

• Read-only. The connector looks up analytics. It can't create, change or delete anything in Hive.
• Your access, nothing more. Each person signs in with their own Hive account, and answers only include data that account can already see.
• What's covered. INA event data: event totals, per-event performance, viewer sites and networks, viewer devices and browsers, CDN and peer-to-peer delivery, and playback reliability and video quality. Per-viewer and per-stream diagnostic data is not available through the connector.
• Your assistant's terms apply. Results are passed to the assistant you use and handled under your organization's agreement with that provider.
• Disconnect any time. Remove the connector from your assistant's settings to revoke access immediately.

TroubleshootingCopy Anchor Link