getting started

How to start a crawl

Enter a URL, pick a crawl mode, preview the scope, and run your first workspace crawl in Silkra.

Start screen

After sign-in you land on the start screen. Enter a URL, reopen a saved workspace from the rail, or upload a URL list. The helper chat on the start screen can answer setup questions.

Crawl modes

Silkra auto-detects mode from your URL when possible.

ModeDiscoveryBest for
SpiderFollows links from a starting URLFull site audits, structure analysis
SitemapReads URLs from sitemap.xmlKnown URL lists, indexed page coverage
ListOnly URLs you provide (CSV or paste)Specific pages, competitor URL sets

Spider vs preview scope

Toggle on the start screen before entering a URL.

ActionWhen to use
SpiderStart crawling from your URL with scope options below
Preview scopeMap folders or sitemaps first, then narrow before crawling

Scope preview

After entering a URL in preview mode, Silkra maps folders from the sitemap, homepage links, or robots.txt. Click folder pills to include or exclude sections. For sitemap indexes, pick child sitemaps. If no map is found, enter folder paths manually. Unreachable sites show a no-map fallback.

Spider scope

Applies to spider mode only. Subfolder scope auto-applies when you start from a path like example.com/blog. Defaults live in Settings → Spider.

ScopeFollowsExample
DomainAll pages on the domain, including subdomainsexample.com → entire site
SubfolderOnly the starting path and belowexample.com/blog → /blog/*
SubdomainOnly the same subdomainblog.example.com → not shop.example.com
ExactSingle URL onlyNo link discovery

Max pages and URL rules

Set max pages on the start screen before you crawl. Expand Refine scope to add include or exclude URL patterns. Plan limits cap the maximum per run. Spider settings (⌘J → Spider) hold defaults for rendering, politeness, and robots.txt.

Spider settings guide →

URL list uploads

Switch to list mode by uploading a CSV of URLs or pasting a list. Preview shows the URL count before you start. Only listed URLs are crawled.

During the crawl

While crawling, the calm crawl view replaces the table. It shows live progress, recent URLs, and a checklist for scrape and semantic map lanes. You can pause, stop, or cancel from the title bar.

After the crawl

When intelligence finishes, the crawl table fills in. Map and Search unlock in the view switcher. The sidebar opens automatically for new workspace crawls.

Semantic map guide →

Versions and re-crawl

Each crawl run is a version inside the workspace. Switch versions from the title bar version selector. Start a new crawl from the title bar or ⌘N to add another version without losing prior runs.