scientific-critical-thinking

Evaluate scientific claims, study designs, biases, and evidence quality using GRADE and Cochrane frameworks.

Updated Sep 2, 2026
One-click install
npx skills add https://github.com/ricfulop/cba-agentic-engineering-bootstrap --skill scientific-critical-thinking-ricfulop
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: scientific-critical-thinking
Source: https://github.com/ricfulop/cba-agentic-engineering-bootstrap/tree/main/skills/scientific-critical-thinking
Command: npx skills add https://github.com/ricfulop/cba-agentic-engineering-bootstrap --skill scientific-critical-thinking-ricfulop

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires requests, python-dotenv, and includes scripts (resource) and references (resource) components.

What problem does it solve? Researchers and reviewers need a systematic way to judge whether scientific claims are supported by evidence, but informal reading often misses methodological flaws, biases, and statistical errors. This Skill provides structured frameworks for critiquing methodology, detecting bias, evaluating statistics, and grading evidence quality. ## Core Features & Use Cases - Methodology and Design Critique: Assess internal, external, construct, and statistical conclusion validity, plus randomization, blinding, and measurement quality. - Bias and Fallacy Detection: Identify cognitive, selection, measurement, and analysis biases (p-hacking, HARKing, publication bias) and name logical fallacies in scientific arguments. - Evidence Grading: Apply GRADE, Cochrane RoB 2, ROBINS-I, and evidence hierarchy frameworks to rate confidence in findings. - Use Case: When reviewing a preprint claiming a causal effect from an observational study, use this Skill to flag confounding, check power and multiple-comparison handling, and recommend appropriately hedged conclusions. ## Quick Start Ask the assistant to critically evaluate the methodology, biases, and evidence quality of the attached research paper using the scientific-critical-thinking skill.

Frequently Asked Questions about scientific-critical-thinking

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I evaluate the quality of evidence in a research paper?

Use the GRADE framework: start from the study design, then downgrade for risk of bias, inconsistency, indirectness, imprecision, or publication bias, and upgrade for large effects or dose-response relationships. The skill's evidence_hierarchy reference walks through each criterion.

How to detect p-hacking and HARKing in a study?

Check for preregistration, compare published outcomes against registered ones, look for p-values clustered just below 0.05, and verify whether hypotheses were stated a priori. The common_biases and statistical_pitfalls references list detection and mitigation strategies.

What is the difference between GRADE and Cochrane Risk of Bias?

Cochrane RoB 2 assesses bias within individual randomized trials, while GRADE rates overall confidence in a body of evidence across studies. The skill applies RoB 2 for single RCTs and GRADE when synthesizing multiple findings.

Can this skill generate diagrams for evidence frameworks?

Yes, optionally. When explicitly requested, it can call the scientific-schematics skill to generate GRADE flowcharts or bias decision trees, which requires an OPENROUTER_API_KEY and sends prompts to the OpenRouter API.

When should I use this skill instead of writing a peer review directly?

Use it for analyzing evidence quality, identifying methodological flaws, and teaching critical appraisal. For drafting formal peer review reports with structured reviewer comments, the skill documentation directs you to the peer-review skill instead.