# Analyze Critically

> Stress-test a claim, plan, diagnosis, or design by writing rejection conditions before interpreting the evidence. Use when the user asks to critique, steelman, red-team, find the kill criterion, or what would change your mind. Do not use for a domain go/no-go review, a causal effect estimate, a resolvable forecast, or ordinary implementation.

- Skill: `sylphxai/analyze-critically` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add sylphxai/analyze-critically`
- Raw SKILL.md: https://api.skillmd.com/api/skills/sylphxai/analyze-critically/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: SylphxAI (https://skillmd.com/u/sylphxai)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/sylphxai/analyze-critically

---


# Analyze Critically

Write rejection conditions before interpreting ambiguous results. Steelman the strongest case against the leading view. Separate observations from inferences.

A generic critical-thinking attitude is not this job. Competing explanations that make different predictions beat a fallacy list. Base rates, incentives, and selection effects move the ranking before extra reading. Buy only the next observation that can kill a live hypothesis.

Open [competing hypotheses](references/competing-hypotheses.md) when the case needs alternative generation, value-of-information, or ACH limits.

Use `review-domain` for a named go/no-go. Use `analyze-causal-inference` when the question is whether a change caused an outcome. Use `forecast-with-calibration` for a resolvable future event. Use `record-structured-deliberation` when several participants must contribute.