AI Dev Radar

Trust

Methodology and scoring

Public documentation on how we prioritize signal over hype: production weights, data confidence and a unified change taxonomy.

v1.0.0Effective from: 2026-08-25

What changed in this version

Initial release: public Radar Score docs, data confidence, fail-closed Breakout, and unified signal taxonomy.

Production scoring weights

These weights mirror deployed TypeScript constants and SQL. Each published version stores a snapshot in the database.

AI Tools — Radar Score

usefulness30
adoption20
updates15
priceValue15
maturity10
community5
documentation5
Total100

Catalog (repos, skills, MCP, LLM)

trend40
developerValue45
antiHype15
Total100

News — Impact Score

developerRelevance35
sourceAuthority25
technicalImpact25
novelty15
Total100

HN community bonus (cap 10 pts)

YouTube

Video Impact and Radar Score use normalized weights (0–1) on views/h, engagement, channel and recency.

impact=1.00 · radar=1.00

Explainable signals

Each change feed row includes taxonomy type, magnitude, source and verification level.

View changes feed