Skip to main content
OpenKol is a verification layer. If it isn’t consistent, it isn’t useful. Reliability comes from three things:
  1. deterministic scoring,
  2. predictable data handling, and
  3. transparent behavior when inputs are missing.

​
Deterministic scoring

Scores are computed from normalized inputs using pure functions.
  • same post + same timestamp + same market data ⇒ same result
  • score components are capped/clamped so one metric can’t dominate everything
  • the scoring system is documented so users can understand what drove the result
Start here: Scoring overview

​
Operational reliability

OpenKol is designed for real-world usage:
  • caching around expensive lookups (posts, candles, analysis snapshots)
  • persistence of analysis snapshots so permalinks remain stable
  • health checks and guardrails to avoid partial or misleading results

​
User trust

Trust is earned by being explicit:
  • OpenKol uses public data (social metadata + market data)
  • results are reproducible and shareable
  • limitations and edge cases are acknowledged rather than hidden
See also: Why trust OpenKol