Agent Analytics

This Skill addresses the pain points product teams face in data analysis, such as frequently switching between tools, manually configuring event tracking, and struggling to quickly gain user behavior insights. In practice, product managers and growth leads often spend significant time on SQL queries, dashboard building, and experiment setup, rather than focusing on decision-making itself. Agent Analytics lets AI agents directly handle these operations, enabling a full workflow—from event tracking setup to growth recommendations—through natural language instructions.

Usage is straightforward: simply run npx @agent-analytics/cli login in your terminal to complete browser authorization, after which you can execute product analysis tasks using natural language commands. For example, you can instruct the AI agent to "view the sign-up conversion funnel for the last 7 days" or "analyze the 7-day retention trend for new users." It will automatically invoke CLI commands to fetch data and generate reports. Supported features include: creating sites, querying events, analyzing funnels, retention, paths, experiments, user context, and even recommending minimal growth actions based on data.

It is ideal for teams or individuals who need to quickly obtain product data insights, especially developers already using AI coding assistants (like Claude Code or Cursor). Product managers, growth leads, and data analysts can complete analyses via conversation directly within their development environment, without leaving the terminal or switching to third-party analytics platforms. For SaaS product teams that frequently run A/B tests or conduct user behavior diagnostics, this Skill can significantly boost analysis efficiency.

We recommend integrating this tool when your team consistently uses an AI coding assistant and ensuring all members are familiar with basic CLI operations. Note that it primarily relies on browser authorization for login and does not support direct API key usage; the free tier has read limits, and you'll need to upgrade to the Pro version when encountering a PRO_REQUIRED error. Also, all analysis commands must be executed via the fixed CLI commands—do not attempt to use curl or call the API directly.