Confidence Scoring

SOP for assigning calibrated confidence scores to causal claims based on evidence quality and quantity.

yogsoth-ai Updated 512 repo stars

File contents

Confidence Scoring

Assign calibrated confidence scores to causal claims.

Tool

CC file edit

Protocol

  1. Count supporting evidence (weighted by evidence hierarchy)
  2. Count contradicting evidence (weighted similarly)
  3. Calculate net confidence: support_weight / (support_weight + contradict_weight)
  4. Adjust for mechanism plausibility (+0.1 if strong mechanism, -0.1 if no mechanism)
  5. Update claim page frontmatter confidence field

HARD-GATE

Yield

Returns: { claim: string, confidence: number, evidence_count: number }

yogsoth-ai/knowledge-structuring commit d258d9f6da

Frequently asked questions

npx get-skill add yogsoth-ai/confidence-scoring-2