# Scientific Method

> Applies empirical inquiry through Observation, Hypothesis Formation, Experimentation, Analysis, and Conclusion.

- Skill: `felipecustodio/scientific-method` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add felipecustodio/scientific-method`
- Raw SKILL.md: https://api.skillmd.com/api/skills/felipecustodio/scientific-method/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- License: MIT
- Author: felipecustodio (https://skillmd.com/u/felipecustodio)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/felipecustodio/scientific-method

---


## Overview & Purpose

The Scientific Method is a step-by-step method for empirical investigation. It transforms speculative problem solving into structured hypothesis testing, ensuring that conclusions are grounded in verifiable evidence and reproducible reasoning.

## When to Use

- **Root Cause Analysis**: Investigating unexplained bugs, performance regressions, or test failures.
- **Empirical Validation**: Testing technical assumptions against experimental data or logs.
- **Scientific & Research Tasks**: Systematic exploration of unknown phenomena or complex data sets.

## Execution Workflow

1. **Observation**: Record objective facts, error logs, or anomalous behavior without bias.
2. **Hypothesis**: create testable, falsifiable hypotheses explaining the observation.
3. **Experiment Setup**: Design a controlled test to prove or disprove the primary hypothesis.
4. **Data Collection & Analysis**: Execute the experiment, measure results, and compare against expectations.
5. **Conclusion**: Accept, modify, or reject the hypothesis based strictly on evidence.

## Expected Output Contract

```markdown
### Scientific Investigation
- **Observation**: [Factual anomaly / error]
- **Hypothesis**: [Falsifiable explanation]
- **Experiment**: [Validation step]
- **Results**: [Measured data]
- **Conclusion**: [Confirmed / Rejected]
```

## Scripts

- `scripts/scientific_method.py` - Deterministic evaluation, state validation, and CLI tool for scientific-method.