bhuntr-event-monitor

Imported from Wang200935/openclaw-skills (misc/misc/misc/misc/misc/misc/misc/misc/misc/misc/bhuntr-event-monitor/SKILL.md). Install upstream with npx skills add Wang200935/openclaw-skills --skill bhuntr-event-monitor. Copyright stays with the author.

BHuntr Event Monitor

Purpose

Continuously scan https://bhuntr.com/tw every10 minutes, deeply fetch each event subpage, recursively crawl all bhuntr.com subpages (no depth limit, de-duplicated), extract structured fields, filter information/tech-related events, store results, and produce a daily07:00 report for the user.

Tools used

• Python3
• requests (HTTP fetch)
• BeautifulSoup (bs4) (HTML parsing; fallback to regex if unavailable)
• OpenClaw cron (scheduled runs)

Data locations

• Raw+filtered data: C:\Users\wang\.openclaw\workspace_second\data\bhuntr\events.json
• Daily summary output: C:\Users\wang\.openclaw\workspace_second\data\bhuntr\daily_summary.md
• Logs: C:\Users\wang\.openclaw\workspace_second\data\bhuntr\logs\bhuntr_scrape.log

How to run manually

python C:\Users\wang\.openclaw\skills\bhuntr-event-monitor\scripts\scrape_bhuntr.py
python C:\Users\wang\.openclaw\skills\bhuntr-event-monitor\scripts\summarize_bhuntr.py

Scheduling (cron)

Create two isolated cron jobs:

###1) Every10 minutes: scrape & store

openclaw cron add \
 --name "BHuntr scrape" \
 --cron "*/10 * * * *" \
 --tz "Asia/Taipei" \
 --session isolated \
 --message "Use bhuntr-event-monitor skill to run scrape_bhuntr.py and update events.json" \
 --delivery none

###2) Daily07:00: summarize & notify

openclaw cron add \
 --name "BHuntr daily brief" \
 --cron "0 7 * * *" \
 --tz "Asia/Taipei" \
 --session isolated \
 --message "Use bhuntr-event-monitor skill to run summarize_bhuntr.py and post today’s info/tech events" \
 --announce

Output schema (events.json)

Each event item includes:

• id: stable hash
• title
• url
• category/tags
• start_date, end_date (if detected)
• organizer (if detected)
• location (if detected)
• summary
• is_info_related: boolean
• fetched_at

Top-level payload includes:

• total, info_total
• page_total and pages (all crawled URLs)

Filtering rule (資訊類)

Mark as 資訊/科技/AI/資安/程式/資料 related if title/summary/tags contain any of:

資訊, 科技, AI, 人工智慧, 資安, 程式, 程式設計, Coding, 軟體, 開發, 電腦,
資料, Data, 機器學習, 深度學習, 雲端, 網路, 黑客松, Hackathon

Notes

• If site uses heavy JS and data is missing, script will attempt to discover API endpoints from JS bundles.
• If no API endpoints are found, script crawls all same-domain pages and extracts event subpages.