Skip to main content
There are three distinct ways to use AI with tinyanalytics: ask questions about your analytics inside the product, connect an outside AI agent to your analytics data, or give these public docs to an assistant that is helping with an integration. Keeping those contexts separate helps the assistant use the right source.

How is AI used inside tinyanalytics?

Use Ask AI inside the dashboard when the question is about your own analytics. It can:
  • answer site-level questions from the same report services as the dashboard;
  • compare the sites you can access across an organization;
  • link you to the report behind an answer; and
  • propose supported goals, dashboards, or annotations, with explicit approval before a change.
tinyanalytics also measures AI as part of your acquisition data:
  • AI traffic shows visits referred by AI assistants and the pages those visitors view.
  • Bot and crawler reports separate automated traffic, including AI crawlers, from human visits.
  • The SQL query builder turns a natural-language question into a query you can review and run against your analytics data.

Give an external AI the right context

Give your coding assistant the relevant tinyanalytics page before asking it to write an integration or troubleshoot a setup. Every page is available as Markdown, through the documentation MCP server, and in the site’s machine-readable index.

Ask a coding assistant to add tinyanalytics using the official documentation context.

Open in Cursor

Copy or open any page in an AI tool

At the top of every page is a menu to take that page to your assistant of choice:
  • Copy page — copy the page as clean Markdown to paste into any chat.
  • Open in ChatGPT / Claude / Perplexity — send the page straight to your assistant with a prompt.
  • View as Markdown — see the raw Markdown behind the page.

Pull the docs into your editor

The same menu connects the docs to your coding tools:
  • MCP — connect the documentation MCP server so an AI agent can search and read these docs directly. This is not the same as the tinyanalytics MCP server: the documentation server serves these public pages, while the tinyanalytics server serves your own analytics data and needs an API key.
  • Cursor / VS Code — open the docs in your editor’s AI assistant to reference them while you build your integration.

Machine-readable index

These docs publish a machine-readable index for large language models, served at the site root:
  • /llms.txt — a structured map of every page.
  • /llms-full.txt — the full docs content in one file.
Point an AI tool at either file to give it the whole documentation set as context.

API reference

Have your AI build against the API.

Ask AI

Ask questions about your own analytics inside tinyanalytics.

MCP server

Connect an AI agent to your analytics data.

Quickstart

Install tinyanalytics in a few minutes.

Integrations

Ask your AI which install guide fits.

How tinyanalytics works

The mental model behind the docs.

Measure AI traffic

Understand referrals and visits from AI assistants.

Ask questions with SQL

Turn a natural-language question into a reviewable query.