Skip to content
#

evidence-scoring

Here are 7 public repositories matching this topic...

Codebase Context gives AI agents understanding of your codebase through semantic code search, team conventions, patterns, and memory, so they use fewer tokens, spend less time, and produce better, more familiar output.

  • Updated Oct 11, 2026
  • TypeScript

Autonomous Real-Time Narrative Intelligence & Liquidity Arbitrage Swarm for Prediction Markets (Polymarket/Kalshi) and the Attention Economy.

  • Updated Oct 3, 2026
  • Python

Add this topic to your repo

To associate your repository with the evidence-scoring topic, visit your repo's landing page and select "manage topics."

Learn more