get_crawl_health

Check the running status of the background crawl engine, including crawl counts, crawl status per site, and stale monitor detection.

Insight

Description

Check the running status of the background crawl engine, including crawl counts, crawl status per site, and stale monitor detection.

Typical user queries:

• "Is the crawl system running normally"
• "How many times has it crawled today"
• "How is the crawl status across each site"
• "Are there any monitors that haven't been updated for a long time"

No parameters required. Directly returns the crawl system health report. Data is scoped to the current user's own monitored products.

Parameters

No parameters.

Returns

Returns crawl system health report, including crawl statistics, monitor status, alert statistics, crawl status per site, and stale monitor detection.

Crawl System Health Report:

📊 Crawl Statistics (past 7 days):
  - Today's crawls: 28
  - Past 7 days crawls: 156

📦 Monitor Status:
  - Total monitors: 12 (10 active)

🔔 Alert Statistics:
  - Today's alerts: 2

🌏 Crawl Status Per Site (past 7 days):
  - MY: 74 crawls, 3 monitors
  - SG: 82 crawls, 5 monitors, 1 with no data
  - TH: 0 crawls, 1 monitor, 1 with no data
  - VN: 0 crawls, 1 monitor, 1 with no data

⚠️ Monitors with no crawl data in the past 7 days (3):
  - Samsung S24 Ultra (SG)
  - Xiaomi Redmi Note 13 (TH)
  - Realme C55 (VN)

⚠️ No crawl records today, crawl engine may not be running. Please check the Crawler extension.

Data Isolation Note

All data returned by this tool is isolated per user, containing only crawl statistics for the current user's own monitored products, and will not display other users' data.

Conversation Examples

• "Is the crawl system running normally"
• "How many times has it crawled today"
• "How is the crawl status across each site"
• "Are there any monitors that haven't been updated for a long time"

API Endpoints

• GET /user/dashboard — Get today's/7-day crawl counts, alert counts, total monitors (isolated by user_id)
• GET /price/summary?days=7 — Get crawl counts per monitor, aggregated by site (isolated by user_id)

Why are telemetry endpoints no longer used?

The previous implementation used /telemetry/summary and /telemetry/site-stats, but crawl events were recorded by the Crawler extension as the reporter, causing sellers querying via MCP to not see crawl data for their own products. Changed to statistics from the PriceHistory table, with data strictly isolated by user_id, ensuring sellers see crawl data for their own monitored products only.