Skip to main content
This page describes how Linkup filters unsafe content and evaluates source quality during indexing and retrieval.

Default excluded and restricted categories

High-risk categories are excluded or blocked by default, including:
  • Malware
  • Phishing
  • Spyware
  • DNS tunneling
  • Potentially unwanted software
  • Child abuse
Questionable content categories are also restricted. Examples include:
  • Hacking-related content
  • Drug-related content
  • Unreliable information patterns
  • Profanity-heavy pages
  • Adult-themed content

Quality scoring

Linkup applies a quality scoring model to evaluate candidate pages during indexing and retrieval. Higher-quality sources are prioritized; low-trust content is de-prioritized.

Crawling safeguards

  • AI-powered filtering blocks malware, phishing, and other malicious sources before indexing or serving
  • Ranking and curation controls reduce exposure to sensitive or prohibited content
  • High-risk sites can be blocked through filtering and DNS security controls
  • Domain exclusions can be applied via the excludeDomains API parameter
  • Crawling respects robots.txt and does not circumvent CAPTCHAs or access controls
  • Only publicly available content is indexed; content behind authentication, paywalls, or registration is not accessed

Book a call

Discuss enterprise index controls and custom filtering requirements.

Enterprise index controls

Under enterprise agreements, additional controls can be configured, including:
  • Blocking of specific domains
  • Blocking of additional content categories
  • Fastlane inclusion of approved sources not yet present in the index
  • Custom refresh rates for selected sources
  • Customer-specific ranking bias (for example, geographic relevance)
Contact contact@linkup.so for enterprise index configuration.