@cryptotaxi247 / SquadScope
Code Commits Issues Pulls Sessions CI/CD
main
main
md 1.54 KB
View source Raw

Skill: Noise Classification

Purpose

Identify and filter coordinated manipulation patterns in GitHub crawl data before editorial analysis, preventing inflated repos from contaminating trend signals.

Detection Heuristics

Star-Farming (W22 pattern)

  • Trigger: ≥3 repos landing within ±10 stars of each other in the same crawl window
  • Confirm: Zero or near-zero fork count on repos with 400+ stars
  • Context: Often clustered in game-cheat, software-unlock, or AI-branded categories
  • Action: Flag as noise:star-farm, exclude from trend aggregation

Fork-Inflation (W23 pattern)

  • Trigger: Fork/star ratio > 10:1 on repos < 30 days old
  • Confirm: Keyword-stuffed descriptions, bot-pattern naming conventions
  • Context: Polymarket, crypto, and automated-trading verticals
  • Action: Flag as noise:fork-inflation, exclude from trend aggregation

Creation-Timestamp Clustering

  • Trigger: ≥5 thematically similar repos created within a 60-minute window
  • Confirm: Similar description templates, near-identical READMEs
  • Action: Flag cluster as noise:coordinated-creation

Evolution Expectation

Manipulation vectors rotate weekly. When a new metric vector appears (beyond stars and forks), add a detection rule here. Watch for: issue-count inflation, sponsor-badge gaming, discussion-count manipulation.

Integration

  • Pre-filter step before editorial scoring
  • Flagged repos should be noted in analysis frontmatter for retrospective audit
  • Noise counts feed the weekly brief's Signal/Noise section