🕸️ httrack

Mirror any website to local disk with HTTrack.

HTTrack is a mature open-source website copier. This skill turns it into an agent-friendly task: give it a URL and it returns a complete offline copy of the site — HTML, images, stylesheets, and links — that can be browsed with no network connection.

What to use it for

• Archiving a page/site for offline reading or as evidence
• Saving documentation, references, or course material before it changes
• Building a local corpus for later search/analysis
• Snapshotting a site at a point in time

Requirements

• httrack binary. Debian/Ubuntu: sudo apt install httrack
• Outbound network access to the site(s) you mirror

Quick start

# full mirror of a site, 2 links deep, 2 parallel connections
httrack "https://example.com" -O ./mirror -r2 -c2

# single page only
httrack "https://example.com/page.html" -O ./snapshot -r0

# use the safe wrapper included with this skill
./mirror.sh "https://example.com" ./mirror 2

Common flags

Flag	Meaning
-O DIR	output directory
-rN	recursion depth (0 = single page)
-cN	number of parallel connections
--robots=1	obey robots.txt
-%v	verbose progress
-i	continue an interrupted mirror
-Y	update an existing mirror (incremental)
-A "*.pdf,*.zip"	only fetch certain file types
-F "user-agent"	set a custom user agent

Safety & legality

• Only mirror sites you are authorized to archive.
• Respect robots.txt and the site's terms of service.
• Mirrored content may be copyrighted — do not redistribute it.
• Keep connection count and depth low to avoid hammering servers.
• Review downloaded files: a mirror can contain scripts, cookies, or pages that were not intended for you.

Files

• SKILL.md — this guide
• mirror.sh — safe wrapper with sane defaults
• README.md — permissions, security & privacy, verification hash