Web Monitor

通过 Python 脚本把指定网页抓取并保存为快照，比较前后内容差异以检测更新；可用 CSS 选择器只监控页面局部（如价格、文章列表），支持文本或 JSON 形式输出，数据默认存在用户目录下的 web-monitor 目录中。

corbin-breton

热度 79.99

最近更新 2026-05-11 17:36:35

分类：自动化依赖：代码执行 / 读写文件应用场景：工作流 / 自动化脚本 / 研究

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add keats-web-monitor

Metadata

name: web-monitor
description: Monitor web pages for content changes and get alerts. Track URLs, detect updates, view diffs. Use when asked to watch a website, track changes on a page, monitor for new posts/content, set up page change alerts, or check if a site has been updated. Supports CSS selectors for targeted monitoring.

Web Monitor

Track web pages for changes. Stores snapshots, computes diffs, supports CSS selectors.

Triggers

Activate this skill when the user wants to:

Watch a website for content changes
Track price changes, new posts, or updated content on a page
Set up page change alerts or notifications
Check if a site has been updated since last visit
Monitor a specific section of a page (with CSS selector)

NOT For

Real-time alerting on dynamic topics — use the topic-monitor skill for news/topic monitoring with importance scoring
Full page archival — this tracks diffs, not full crawl archives
JavaScript-heavy SPAs — the monitor fetches raw HTML; dynamic content rendered by JS may not be captured accurately
Monitoring APIs or webhooks — use cron + exec for structured data endpoints

Quick Start

# Add a URL to watch
uv run --with beautifulsoup4 python scripts/monitor.py add "https://example.com" --name "Example"

# Add with CSS selector (monitor specific section)
uv run --with beautifulsoup4 python scripts/monitor.py add "https://example.com/pricing" -n "Pricing" -s ".pricing-table"

# Check all watched URLs for changes
uv run --with beautifulsoup4 python scripts/monitor.py check

# Check one specific URL
uv run --with beautifulsoup4 python scripts/monitor.py check "Example"

# List watched URLs
uv run --with beautifulsoup4 python scripts/monitor.py list

# View last diff
uv run --with beautifulsoup4 python scripts/monitor.py diff "Example"

# View current snapshot
uv run --with beautifulsoup4 python scripts/monitor.py snapshot "Example" --lines 50

# Remove
uv run --with beautifulsoup4 python scripts/monitor.py remove "Example"

Commands

Command	Args	Description
`add`	`<url> [-n name] [-s selector]`	Add URL to watch, take initial snapshot
`remove`	`<url-or-name>`	Stop watching a URL
`list`	`[-f json]`	List all watched URLs with stats
`check`	`[url-or-name] [-f json]`	Check for changes (all or one)
`diff`	`<url-or-name>`	Show last recorded diff
`snapshot`	`<url-or-name> [-l lines]`	Show current snapshot

Output Symbols

🔔 CHANGED — page content changed (shows diff preview)
✅ No changes
📸 Initial snapshot taken
❌ Error fetching

Data

Stored in ~/.web-monitor/ (override with WEB_MONITOR_DIR env var). For OpenClaw workspace integration, set WEB_MONITOR_DIR=$WORKSPACE/data/web-monitor (where $WORKSPACE is your agent workspace path).

watches.json — watch list config
snapshots/ — stored page content + diffs

Tips

Use --selector to monitor specific elements (prices, article lists, etc.)
Use --format json for programmatic checking (heartbeat integration)
CSS selectors require beautifulsoup4 (included via --with flag)
Text is normalized to reduce noise from timestamps, whitespace, ads