菜单
Use when auditing webpages or URL lists for broken links, redirects, crawl health, HTTP status failures, timeouts, or link-report generation.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | link-checker |
| description | Use when auditing webpages or URL lists for broken links, redirects, crawl health, HTTP status failures, timeouts, or link-report generation. |
Broken link finder that crawls web pages and checks all hyperlinks for HTTP errors (404 Not Found), redirects (301/302), timeouts, and connection failures. Generates detailed reports with status codes, response times, and link locations. Supports recursive crawling with depth control, domain filtering, and multiple output formats. Uses Python3 urllib — no external dependencies. Essential for website maintenance, SEO hygiene, and quality assurance.
Powered by BytesAgain | bytesagain.com | hello@bytesagain.com
| Command | Description |
|---|---|
check <url> | Check all links on a single page |
crawl <url> | Recursively crawl and check links |
report <url> | Generate a full HTML report |
broken <url> | Show only broken links |
redirects <url> | Show only redirected links |
external <url> | Check only external links |
internal <url> | Check only internal links |
batch <file> | Check links on multiple pages |
summary <url> | Quick summary with counts |
--depth <n> — Crawl depth for recursive mode (default: 1)--timeout <seconds> — Request timeout per link (default: 10)--format text|json|csv|html — Output format (default: text)--output <file> — Save report to file--concurrent <n> — Max concurrent checks (default: 5)--internal-only — Only check internal links--external-only — Only check external links--exclude <pattern> — Exclude URLs matching pattern--verbose — Show all links including OK ones--quiet — Only show broken links--user-agent <string> — Custom User-Agent header# Check all links on a page
bash scripts/main.sh check https://example.com
# Find only broken links
bash scripts/main.sh broken https://example.com --format json
# Recursive crawl with depth limit
bash scripts/main.sh crawl https://example.com --depth 2 --internal-only
# Generate HTML report
bash scripts/main.sh report https://example.com --output link-report.html
# Check external links only with custom timeout
bash scripts/main.sh external https://example.com --timeout 15
# Quick summary
bash scripts/main.sh summary https://example.com
| Category | Status Codes | Description |
|---|---|---|
| ✅ OK | 200, 204 | Link is working |
| 🔄 Redirect | 301, 302, 307, 308 | Link redirects |
| ❌ Broken | 404, 410 | Link is dead |
| ⚠️ Error | 403, 500, 502, 503 | Server error |
| ⏱️ Timeout | — | Connection timed out |
| 🚫 Failed | — | Connection refused/DNS failure |
Route ACP-backed agent requests through the bundled acpx backend.
Apply Vercel's React and Next.js performance optimization rules when auditing or refactoring existing React/Next.js code for waterfalls, bundle size, hydration issues, render cost, or measurable runtime performance problems.
Use when building or improving browser-facing UI where visual hierarchy, responsive layout, interaction design, motion, media, copy, or product polish is the main challenge.
将图文/视频内容发布到小红书（XHS），并支持登录检查、二维码登录辅助、搜索笔记、获取详情、评论/回复、读取内容数据和多账号/多端口运行。使用 Python/CDP 控制面。
OpenProse VM skill pack. Activate on any `prose` command, .prose files, or OpenProse mentions; orchestrates multi-agent workflows.
Use when the user asks to find, compare, install, vet, or create skills, including Chinese requests like 技能, 找技能, 安装 skill, find-skill, find-skills, or install skill.