# Evidence Weighing

> Tactic for assessing the strength and relevance of evidence for causal claims — distinguishes correlation from causation.

- Skill: `yogsoth-ai/evidence-weighing-2` (Agent Skill)
- Install (CLI): `npx get-skill add yogsoth-ai/evidence-weighing-2`
- Raw SKILL.md: https://api.skillmd.com/api/skills/yogsoth-ai/evidence-weighing-2/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: yogsoth-ai (https://skillmd.com/u/yogsoth-ai)
- Updated: 2026-09-21
- Page: https://skillmd.com/skills/yogsoth-ai/evidence-weighing-2

---


# Evidence Weighing

Assess the strength of evidence for causal claims. Not all evidence is equal — RCTs > observational studies > expert opinion > anecdote.

## Available SOPs

- evidence-linking — connect evidence to claims
- confidence-scoring — assign calibrated confidence
- contradiction-flagging — identify conflicting evidence

## Guiding Principles

- **Hierarchy of evidence.** Interventional > observational > theoretical > anecdotal.
- **Effect size matters.** A statistically significant but tiny effect is weak evidence for practical causation.
- **Replication is king.** Single studies are suggestive; replicated findings are convincing.
- **Confounders weaken.** Evidence from studies with uncontrolled confounders gets lower weight.
- **Mechanism strengthens.** Evidence backed by a plausible mechanism gets higher weight.

## Minimum Yield

<HARD-GATE>
≥2 evidence assessments with explicit strength ratings per invocation.
</HARD-GATE>

