Market Screen

Rank multiple A-share candidates for: $ARGUMENTS

Overview

Implementation status: code-backed
Local entry script: <bundle-root>/market-screen/run.py
Primary purpose: batch-score a candidate universe and return the strongest names first
Research layer: universe formation and comparative ranking (Stage 1: Strategy Design & Hypothesis Formation - Universe definition, Stage 4: Feature Engineering & Signal Construction - Cross-sectional ranking)
Workflow stages: stage 1 Universe Formation and stage 4 Feature Engineering & Signal Construction for breadth-first triage
Local executor guarantee: apply the shared market score to each candidate, sort valid names, and list skipped entries when partial failures occur