# analyga

> Revenue analytics for AI-native SaaS. Board-ready numbers organised around your growth funnel — and the only dashboard that counts agent (MCP/AI) traffic as a channel. A dashboard for humans; the `tga` CLI and an MCP server for your agents.

analyga is a privacy-friendly web analytics platform (an Umami-based fork) built for the revenue owner at an AI-native SaaS. It shows what pays — stats, funnels, retention, journeys, attribution, revenue and network rollups — organised around the growth funnel, in language a board reads, and it is the only analytics that counts agent traffic (MCP and AI calls) alongside human traffic. The same analytics are exposed through three surfaces: a web dashboard, the `tga` command-line tool, and a Model Context Protocol (MCP) server. Agents can self-signup, send server-side events, mint ingest keys, and read their own schema without ever opening a browser.

## Get started
- Install the CLI / MCP server: `npx analyga` (npm package: `analyga`, binary: `tga`)
- Onboarding is keyless: an agent can self-signup via the `analyga_signup` tool, then use its `tga_...` API key.
- Remote MCP endpoint: https://mcp.analyga.com/sse (Authorization: `Bearer tga_...`)

## Key features
- Query stats, funnels, retention, journeys, attribution and revenue from a terminal or MCP client
- Server-side event ingest with per-website ingest keys
- Network rollups across multiple websites
- Structured JSON on every command (`--json`) so agents can branch on the numbers

## Links
- Homepage: https://analyga.com
- npm: https://www.npmjs.com/package/analyga
- MCP registry name: io.github.jyswee/analyga