Lcontext

The product analytics agent for your AI coding tools. Lcontext studies how users navigate your product, where they engage, hesitate, and drop off, and delivers behavioral context via MCP.

Works with: Claude Code, Claude Desktop, Cursor, Windsurf, Cline, and any MCP-compatible AI coding tool.

Setup

1. Sign Up

Create a free account at lcontext.com. This gives you access to the dashboard, your tracking script tag, and your API key.

2. Add the Tracking Script to Your App

The tracking script collects behavioral data from your app so Lcontext can study how users navigate your product.

Add this to your app's HTML:

<script src="https://lcontext.com/it.js?iTag=YOUR_SITE_TAG" defer></script>

Replace YOUR_SITE_TAG with the tag from your app settings in the dashboard.

• Page views - URL, title, referrer
• Clicks - Buttons, links, interactive elements (label text only, no personal data)
• Form submissions - Form metadata (no field values captured)
• Scroll depth - How far users scroll (25% increments)
• Device metadata - Screen size, browser language, viewport

Privacy-focused: No passwords, input values, or personal information collected. ~4KB minified.

Associate the anonymous visitor with your application's user ID:

lcontext('setProfileId', 'user-123');

Track custom events:

lcontext('trackEvent', 'purchase', { productId: 'abc', amount: 99.99 });

3. Connect Your AI Coding Tool

Find your API key in Settings > API Access in the dashboard, then set up your tool:

Run the install script - it will prompt for your API key and configure Claude automatically:

curl -fsSL https://lcontext.com/api/cli/install | bash

Then restart Claude.

Option 1: One-Click Install

Install Lcontext in Cursor

After clicking, replace your-api-key-here with your actual API key in Cursor settings.

Option 2: Manual Setup

First, install the binary:

curl -fsSL https://lcontext.com/api/cli/install | bash

Then open Cursor Settings (Cmd/Ctrl + Shift + J), and add the server:

{
  "lcontext": {
    "command": "lcontext",
    "env": {
      "LCONTEXT_API_KEY": "your-api-key-here"
    }
  }
}

First, install the binary:

curl -fsSL https://lcontext.com/api/cli/install | bash

Then open Windsurf and navigate to Cascade > Configure > MCP Servers (or click the hammer icon), then click "Add Server" and select "Add custom server". Add this configuration:

{
  "lcontext": {
    "command": "lcontext",
    "env": {
      "LCONTEXT_API_KEY": "your-api-key-here"
    }
  }
}

First, install the binary:

curl -fsSL https://lcontext.com/api/cli/install | bash

Then in VS Code with Cline installed:

1. Open Cline settings (click gear icon in Cline panel)
2. Navigate to MCP Servers section
3. Add a new server:

{
  "lcontext": {
    "command": "lcontext",
    "env": {
      "LCONTEXT_API_KEY": "your-api-key-here"
    }
  }
}

First, install the binary:

curl -fsSL https://lcontext.com/api/cli/install | bash

Lcontext works with any tool that supports the Model Context Protocol. The general configuration pattern is:

• Command: lcontext
• Environment Variable: LCONTEXT_API_KEY=your-api-key-here

Refer to your tool's documentation for MCP server configuration.

Alternative Installation Methods

The install script used above (curl -fsSL https://lcontext.com/api/cli/install | bash) is the recommended approach. If you need an alternative:

Download the binary for your platform from the latest release:

Platform	Binary
macOS (Apple Silicon)	lcontext-macos-arm64
macOS (Intel)	lcontext-macos-x64
Linux (x64)	lcontext-linux-x64
Linux (ARM64)	lcontext-linux-arm64
Windows (x64)	lcontext-windows-x64.exe

If you have Node.js installed:

npm install -g lcontext-mcp

git clone https://github.com/lcontext/lcontext.git
cd lcontext
npm install
npm run build

Available Tools

get_page_context

Get comprehensive analytics context for a page including stats, visitor metrics, Web Vitals (LCP, FCP, FID, CLS), and all interactive elements with their engagement data.

Parameters:

• path (required): The page path (e.g., /products, /checkout)
• startDate (optional): Start date for stats (ISO format)
• endDate (optional): End date for stats (ISO format)
• periodType (optional): day or week (default: day)

Example:

Get the analytics context for the /checkout page for the last 7 days

list_pages

List all tracked pages for your app.

Parameters:

• limit (optional): Maximum pages to return (default: 50, max: 200)
• search (optional): Filter by path (e.g., /product)

Example:

What pages are being tracked on my app?

get_element_context

Get detailed analytics for a specific interactive element by its label or ID.

Parameters:

• elementLabel (optional): Element's label text or aria-label
• elementId (optional): Element's HTML ID
• pagePath (optional): Filter by page path

Example:

Show me analytics for the "Add to Cart" button

get_app_context

Get application-wide analytics including sessions, visitors, page views, engagement metrics, device breakdown (mobile/tablet/desktop), top browsers, top OS, Web Vitals (LCP, FCP, FID, CLS), and AI-generated insights.

Parameters:

• periodType (optional): day or week (default: day)
• limit (optional): Number of periods to return (default: 7, max: 30)

Example:

Give me an overview of my app's behavioral analytics for the last week

get_visitors

Get a list of visitors with AI-generated profiles, interests, engagement trends, segment assignments, device info (type, browser, OS), and location (city, region, country).

Parameters:

• limit (optional): Maximum visitors to return (default: 20, max: 100)
• offset (optional): Offset for pagination
• segmentId (optional): Filter by segment ID
• search (optional): Search in visitor ID, title, summary, interests, goals, action, evidence
• firstVisitAfter (optional): Filter by first visit date (ISO format)
• firstVisitBefore (optional): Filter by first visit date (ISO format)
• lastVisitAfter (optional): Filter by last visit date (ISO format)
• lastVisitBefore (optional): Filter by last visit date (ISO format)
• engagementTrend (optional): increasing, stable, or decreasing
• overallSentiment (optional): positive, negative, neutral, or mixed

Example:

Show me visitors with increasing engagement trend

get_visitor_detail

Get detailed profile and recent sessions for a specific visitor, including device info (type, browser, OS) and location.

Parameters:

• visitorId (required): The visitor's unique identifier

Example:

Get the full profile for visitor abc123

get_sessions

Get a list of user sessions with AI-generated summaries, titles, sentiment analysis, and device type.

Parameters:

• limit (optional): Maximum sessions to return (default: 20, max: 100)
• offset (optional): Offset for pagination
• visitorId (optional): Filter by visitor ID
• sentiment (optional): positive, negative, or neutral
• startDate (optional): Start date for filtering (ISO format)
• endDate (optional): End date for filtering (ISO format)
• search (optional): Search in session title and description
• minDuration (optional): Minimum session duration in seconds
• maxDuration (optional): Maximum session duration in seconds
• minEventsCount (optional): Minimum number of events
• maxEventsCount (optional): Maximum number of events
• pagePath (optional): Filter sessions that visited a specific page path

Example:

Show me negative sentiment sessions from the last 24 hours

get_session_detail

Get detailed information about a specific session including full event timeline, visitor context, device type, and location (city, region, country).

Parameters:

• sessionId (required): The session's numeric ID

Example:

Show me the details of session 12345

get_user_flows

Get automatically detected user journey patterns showing how users navigate through the application. Each flow represents a common page sequence with engagement metrics and drop-off points.

Parameters:

• limit (optional): Maximum flows to return (default: 10, max: 50)
• category (optional): Filter by category: conversion, exploration, onboarding, support, engagement, other
• minSessions (optional): Minimum session count for a flow to be included
• periodType (optional): day for daily flows, week for weekly aggregated flows

Example:

Show me the most common user journeys on my app

get_analysis

Get a pre-computed daily or weekly analysis report. Contains problem areas, funnel health, session findings, and prioritized recommendations — all referencing specific page paths, element IDs, and session IDs you can investigate with other tools.

Parameters:

• periodType (optional): day for daily report (default), week for weekly report
• date (optional): ISO date string for the period start (e.g., 2026-02-23). Defaults to most recent completed analysis.

Example:

Get the latest daily analysis for my app

Prompts

analytics-guide

A comprehensive guide for analyzing user behavior data with Lcontext tools. Includes step-by-step analysis workflows, decision trees for common scenarios (e.g., "this page isn't converting", "performance feels slow"), and a reference table connecting analytics findings to code investigations.

AI agents can request this prompt to get expert-level guidance on how to use the tools effectively.

Updating

The binary includes a self-update command:

lcontext --update

Environment Variables

Variable	Description	Required
LCONTEXT_API_KEY	Your Lcontext API key	Yes
LCONTEXT_API_URL	API base URL (default: https://lcontext.com)	No

Contributing

See CONTRIBUTING.md for development setup and build instructions.

License

MIT - see LICENSE for details.