Review Monitor

Review Monitor

Monitor and respond to reviews on Google and Yelp with sentiment analysis and response templates.

Overview

This skill helps businesses monitor their online reputation by:

• Tracking reviews from Google and Yelp
• Analyzing sentiment (positive, neutral, negative)
• Suggesting response templates
• Sending alerts for new reviews
• Maintaining review history

Features

• Multi-Platform Support: Monitor Google and Yelp reviews
• Sentiment Analysis: Automatic sentiment detection using keyword analysis
• Response Templates: Pre-built templates for different review types and sentiments
• Alert Notifications: Get notified of new reviews
• Review History: Track all reviews over time
• Statistics: Analytics on review trends and sentiment distribution

Installation

cd skills/review-monitor
npm install
npm run build

Usage

CLI

# Add a business to monitor
npm run cli -- add-business --name "My Restaurant" --platform google --location-id "ChIJ..."

# List monitored businesses
npm run cli -- businesses

# Fetch new reviews
npm run cli -- fetch

# View reviews
npm run cli -- reviews --platform google

# Analyze sentiment
npm run cli -- analyze --review-id 123

# Generate response
npm run cli -- respond --review-id 123

# View statistics
npm run cli -- stats

Library

import { ReviewMonitorSkill } from '@openclaw/review-monitor';

const monitor = new ReviewMonitorSkill();

// Add a business
const business = await monitor.addBusiness({
  name: 'My Restaurant',
  platform: 'google',
  externalId: 'ChIJ...',
  location: 'New York, NY'
});

// Fetch reviews
const reviews = await monitor.fetchReviews(business.id!);

// Analyze sentiment
const sentiment = await monitor.analyzeSentiment(reviews[0].id!);

// Generate response
const response = await monitor.generateResponse(reviews[0].id!, 'grateful');

// Close connection
await monitor.close();

Configuration

The skill stores data in ~/.openclaw/skills/review-monitor/.

Response Templates

Available template types:

• grateful - For positive reviews
• apologetic - For negative reviews
• professional - For neutral reviews
• follow-up - For questions or concerns
• appreciation - General thank you

Sentiment Analysis

Sentiment is determined by:

• Positive keywords (excellent, great, amazing, etc.)
• Negative keywords (terrible, awful, disappointed, etc.)
• Rating correlation
• Context analysis

Scores range from -1.0 (very negative) to +1.0 (very positive).