# DeadLink

> Scan websites and files for broken links with HTTP status details. Use when auditing links, finding broken URLs, validating references.

- Skill: `bytesagain/deadlink` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add bytesagain/deadlink`
- Raw SKILL.md: https://api.skillmd.com/api/skills/bytesagain/deadlink/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: bytesagain (https://skillmd.com/u/bytesagain)
- Updated: 2026-09-09
- Page: https://skillmd.com/skills/bytesagain/deadlink

---


# DeadLink — Dead Link Checker

Check URLs for broken links. Scan individual URLs, files containing links, or crawl websites. Generates reports with HTTP status codes.

## Commands

| Command | Description |
|---------|-------------|
| `check <url>` | Check a single URL — shows HTTP status, redirect target if applicable |
| `scan <file>` | Extract and check all URLs found in a text file |
| `site <url> [depth]` | Crawl a webpage, extract all href/src links, and check each one |
| `report <file>` | Generate a timestamped report file from all URLs in a text file |

## Examples

```bash
# Check a single URL
deadlink check [Technical Reference]
# → 200 OK

# Scan a markdown file for broken links
deadlink scan README.md
# → Extracts all http/https URLs and checks each one

# Crawl a website
deadlink site [Technical Reference] 1
# → Fetches the page, extracts all links, checks each

# Generate a report file
deadlink report bookmarks.html
# → Creates deadlink-report-20240101-120000.txt
```

## Status Codes

- **2xx** — OK (alive)
- **3xx** — Redirect (alive, shows final URL)
- **4xx** — Client error (dead — 404 Not Found, 403 Forbidden, etc.)
- **5xx** — Server error (dead)
- **000** — Connection failed (DNS error, timeout, unreachable)

## Requirements

- `curl` — must be installed and in PATH
- Network access to check URLs

## Notes

- Timeout: 10 seconds per URL (5s connect timeout)
- URLs are extracted using regex pattern matching for `[Technical Reference] and `[Technical Reference] links
- The `site` command does basic HTML link extraction (href and src attributes)
- Reports are saved as plain text files in the current directory