Grade the quality and confidence of evidence from metamaterial literature. Use when you need to assess how reliable extracted design patterns and claims are.
Assign confidence grades (high/medium/low) to extracted design claims using a weighted scoring formula.
When this skill is invoked (typically after synthesis):
Grade each claim from the synthesis using six factors:
Formula: Score = 0.30R + 0.15V + 0.15T + 0.20M + 0.10L + 0.10C
Assign grades: high (≥0.6), medium (0.35–0.6), low (<0.35)
Compute overall evidence quality from the distribution of grades.
Run via Python:
cd D:/Claude && python -c "
from target_to_hypothesis.skills.evidence_grader import grade_evidence
# grades = grade_evidence(paper_records, synthesis)
"
D:/Claude/target_to_hypothesis/skills/evidence_grader.py
D:/Claude/target_to_hypothesis/config/scoring_weights.yaml