Signal Layer

Tracking since May 23, 2026.

Changelog

• Jun 11, 2026
  description

  Signal Layer — Content Signal Analyzer
   
  The internet is filling with template-generated content, SEO farms, and AI-expanded articles that are hard to distinguish from genuine human writing. Signal Layer gives you a fast, transparent, privacy-first tool to evaluate what you are actually reading.
   
  HOW IT WORKS
  ────────────
  Click the Signal Layer icon on any page. The extension analyzes the visible text entirely within your browser — no data leaves your device, no API is called, no server is contacted. In seconds you receive three outputs:
   
  • Human Trace Score (0–100) — A single composite number. Higher scores indicate organically structured text with strong lexical diversity. Lower scores reflect dense template patterns and low linguistic variation.
   
  • Synthetic Content Signal (Low / Medium / High) — Evaluates AI-associated transitional phrase density, sentence rhythm uniformity, and vocabulary repetition.
   
  • Template Repetition Signal (Low / Medium / High) — Detects repeated trigram clusters, affiliate and commission links, and bullet-heavy structural layouts.
   
  A short plain-language explanation accompanies every result.
   
  WHAT MAKES SIGNAL LAYER DIFFERENT
  ──────────────────────────────────
  Most "AI detector" tools send your text to a remote server and return a black-box percentage. Signal Layer does the opposite:
   
  ✓ Zero data collection — nothing is transmitted, stored, or logged
  ✓ No API dependency — works offline, no OpenAI or Anthropic key required
  ✓ No percentage claims — scores reflect statistical text patterns, not authorship verdicts
  ✓ Academic-safe — the scoring model distinguishes formal academic writing from synthetic SEO content, avoiding false positives on structured scholarly text
  ✓ Transparent affiliate detection — separates genuine commission links (Amazon ref=, ShareASale, Awin) from standard analytics trackers (utm_source, utm_campaign)
   
  IMPORTANT NOTICE
  ────────────────
  Signal Layer is a heuristic statistical tool, not an AI authorship detection system. The Human Trace Score measures linguistic repetition and structural uniformity — it is not a definitive proof of AI generation and must not be used as the sole basis for academic, disciplinary, or legal decisions. Full disclaimer: [your privacy policy URL]
   
  PERMISSIONS USED
  ────────────────
  • activeTab — reads the content of the tab you are currently viewing, only when you click "Analyze This Page"
  • scripting — injects the local analysis script into that tab
   
  No other permissions are requested. No background processes run.
   
  IDEAL FOR
  ─────────
  Researchers and academics verifying source quality
  Journalists evaluating content provenance
  Students who want to understand the structure of what they read
  Anyone tired of SEO farms and template-generated content

  Signal Layer — Privacy-First Multilingual Web Quality Analyzer
  
  The internet is filling with template-generated content, SEO farms, and AI-expanded articles. Signal Layer gives you a fast, transparent, privacy-first tool to evaluate what you are actually reading — in 20 languages.
  
  HOW IT WORKS
  ────────────
  Click the Signal Layer icon on any page. The extension analyzes the visible text entirely within your browser — no data leaves your device, no API is called, no server is contacted. In seconds you receive:
  
  • Human Trace Score (0–100) — A single composite number. Higher scores indicate organically structured text. Lower scores reflect dense template patterns and synthetic linguistic rhythm.
  
  • Synthetic Content Signal (Low / Medium / High) — Evaluates AI-associated transitional phrase density, sentence rhythm uniformity, and vocabulary repetition.
  
  • Template Repetition Signal (Low / Medium / High) — Detects repeated phrase clusters, affiliate and commission links, and bullet-heavy structural layouts.
  
  A plain-language explanation accompanies every result.
  
  MULTILINGUAL SUPPORT — 20 LANGUAGES
  ─────────────────────────────────────
  Signal Layer v0.5 introduces two-tier multilingual analysis:
  
  Tier A — Full Linguistic Analysis (7 languages)
  English · Turkish · Spanish · French · German · Portuguese · Italian
  
  Full support includes: phrase density detection, sentence rhythm analysis, lexical diversity (TTR), trigram repetition, and structural pattern analysis. Each language has its own calibrated phrase list and detection thresholds to minimize false positives.
  
  Tier B — Structural Analysis (13 languages)
  Russian · Arabic · Chinese · Japanese · Korean · Hindi · Dutch · Polish
  Vietnamese · Indonesian · Thai · Hebrew · Ukrainian
  
  Structural support includes: DOM analysis, heading density, paragraph fragmentation, affiliate link detection, and template repetition. Works reliably across writing systems where whitespace-based tokenization is less applicable.
  
  The extension automatically detects the page language and applies the correct analysis tier — no configuration needed.
  
  WHAT MAKES SIGNAL LAYER DIFFERENT
  ───────────────────────────────────
  Most tools send your text to a remote server. Signal Layer does the opposite:
  
  ✓ Zero data collection — nothing transmitted, stored, or logged
  ✓ No API dependency — works offline, no subscription key required
  ✓ No authorship claims — scores describe linguistic patterns, not AI origin
  ✓ Academic-safe — scoring model distinguishes formal scholarly writing from synthetic SEO content across languages
  ✓ Transparent affiliate detection — separates genuine commission links from standard analytics trackers
  ✓ Smart caching — revisiting a site shows your previous score instantly
  ✓ Opt-in calibration feedback — voluntary 👍/👎 buttons help improve accuracy
  
  IMPORTANT NOTICE
  ────────────────
  Signal Layer is a heuristic statistical tool, not an AI authorship detection system. The Human Trace Score measures linguistic repetition and structural uniformity — it is not definitive proof of AI generation and must not be used as the sole basis for academic, disciplinary, or legal decisions.
  
  PERMISSIONS USED
  ────────────────
  • activeTab — reads the current tab only when you click "Analyze This Page"
  • scripting — injects the local analysis script into that tab
  • storage — caches analysis results locally on your device (24-hour expiry)
  
  No background processes. No other permissions requested.
  
  IDEAL FOR
  ─────────
  - Researchers and academics verifying source quality across languages
  - Journalists evaluating multilingual content provenance
  - Students assessing the reliability of sources in their native language
  - Anyone reading international web content and wanting a quality signal

• Jun 11, 2026
  permissions

  activeTab, scripting, storage

  activeTab, scripting

• May 29, 2026
  permissions

  activeTab, scripting

  activeTab, scripting, storage

Permissions & access

Permissions

activeTabscripting

Host access

<all_urls>

Screenshots