GTS Ecommerce Price Monitor

Monitor competitor product prices on Shopee, Lazada, Amazon, and others with automated tracking, alerts, reports, and trend analysis for smarter pricing.

E-commerce Competitor Price Monitor

Automatically track competitor product prices across Shopee, Lazada, Amazon, and other platforms. Get notified on price drops, generate weekly comparison reports, and make data-driven pricing decisions.

Quick Start

1. Configure targets: Tell the agent your competitor product URLs and tracking frequency.
2. Run a scan: scan prices now — the agent visits each URL, extracts the price, and logs it.
3. Get a report: price report or compare prices — generates a clean comparison table.
4. Set alerts: alert me when [product] drops below [price] — the agent checks on schedule and notifies you.

Workflow

Step 1: Add Products to Track

Tell the agent in natural language:

"Track the price of this Shopee product: https://shopee.com.my/... every 6 hours" "Add this Lazada link to my price monitor: https://www.lazada.com.my/..." "Monitor these 5 Amazon products daily"

The agent should:

• Accept any e-commerce product URL
• Ask for the desired check frequency (default: 24h)
• Ask for an optional target/alert price
• Store the config in scripts/prices.json

Step 2: Price Scanning

When the user asks to scan or on schedule, the agent:

1. Reads the product list from scripts/prices.json
2. For each URL:
   • Visits the page using web browser
   • Extracts: product name, current price, currency, stock status, seller name
   • Handles platform-specific page structures (see PLATFORMS.md)
3. Logs results to scripts/price_history.jsonl (append-only)
4. Updates latest prices in scripts/prices.json
5. Checks alert conditions — if triggered, notifies the user

Step 3: Reports

On request, generate:

Quick comparison (inline Telegram-friendly):

text

📊 Price Watch — 3 products tracked
───────────────────────────────
Product A  → RM 89.90  ▼ 5% from last week
Product B  → RM 245.00  ▲ 2% (new all-time high)
Product C  → RM 59.90  — No change

Detailed CSV report: Run the Python script scripts/report.py to generate reports/price_report_YYYY-MM-DD.csv with full history.

Step 4: Alerts

The agent supports natural language alert setup:

"Tell me if iPhone 15 drops below RM 4,000" "Alert me when Product C goes on sale" "Notify me if any tracked product drops more than 15%" "Stop alerting me about Product A"

Alerts are stored in scripts/alerts.json. The agent checks them after each scan cycle.

Scripts

scripts/manage.py — Data management CLI

bash

python3 scripts/manage.py products add <url> --name "Product Name" --interval 6 --alert-price 89.90
python3 scripts/manage.py products list
python3 scripts/manage.py products remove <id>
python3 scripts/manage.py alerts add --product <id> --type below --value 100.00
python3 scripts/manage.py alerts list
python3 scripts/manage.py alerts remove <id>
python3 scripts/manage.py history export --days 30

scripts/report.py — Generate reports

bash

python3 scripts/report.py                    # Quick summary (stdout)
python3 scripts/report.py --csv              # Full CSV to reports/
python3 scripts/report.py --product <id>     # Single product history
python3 scripts/report.py --trends           # Price trend analysis

References

• Platform scraping notes: See references/PLATFORMS.md for platform-specific tips (Shopee dynamic loading, Lazada anti-scrape, Amazon structure)
• Configuration: See references/CONFIG.md for advanced setup (custom intervals, notification channels, export formats)

Tips for Best Results

• Shopee: Page loads dynamically. Wait for .product-price or [data-testid="product-price"] elements.
• Lazada: May show region-specific prices. Check URL for .my (Malaysia), .sg (Singapore), etc.
• Amazon: Product prices are in #corePriceDisplay_desktop_feature_div .a-price-whole. Watch for coupon discounts listed separately.
• Frequency: For fast-moving products (electronics), check every 6h. For stable goods (household), daily is fine.
• Freshness: Always verify the extracted price looks reasonable before logging (reject nulls, zeroes, or parsing errors).