Analitoly

AI-аналитик для любого продукта.

Installation

npm (recommended)

npm install -g analitoly
analitoly

Open http://localhost:40000 in your browser.

Docker

docker-compose up -d

From source

git clone <repo-url>
cd analitoly
npm install          # installs all dependencies (root, api, client)
npm run api          # start backend on :40000
npm run client       # start frontend on :30000 (dev mode with hot reload)

Configuration

| Variable | Default | Description | |----------|---------|-------------| | PORT | 40000 | HTTP server port | | ANALITOLY_DATA_DIR | ~/.analitoly | Data directory for config and dashboards | | ANTHROPIC_API_KEY | — | Required for AI chat features | | GRAYLOG_BASE_URL | — | Optional: Graylog integration | | GRAYLOG_USERNAME | — | Optional: Graylog credentials | | GRAYLOG_PASSWORD | — | Optional: Graylog credentials | | GRAFANA_URL | — | Optional: Grafana integration | | GRAFANA_SERVICE_ACCOUNT_TOKEN | — | Optional: Grafana credentials |

CLI flags: --port <number>, --data-dir <path>, --help, --version

Claude Code Integration (MCP)

After installing analitoly globally, add to your Claude Code MCP config:

{
  "mcpServers": {
    "analitoly": {
      "type": "stdio",
      "command": "analitoly-mcp"
    }
  }
}

This gives Claude Code access to your databases, Graylog, and Grafana through MCP tools.

Быстрый старт

Что нужно

• Node.js 20+
• npm

Установка

git clone <repo-url> && cd analitoly
cd client && npm i && cd ..
cd api && npm i && cd ..

Запуск

В двух терминалах:

# терминал 1 — бэкенд
npm run api

# терминал 2 — фронт
npm run client

Открыть http://localhost:30000