Scientific Reasoning
- 17 installs
- 869 repo stars
- Updated June 8, 2026
- beita6969/scienceclaw
scientific-reasoning is a Claude skill that solves mathematical and physical reasoning problems with rigorous step-by-step proofs and verification.
About
This skill solves mathematical and physical reasoning problems with rigorous step-by-step proofs. A user applies it to competition math, applied math, and physics problems that need explicit derivations. It enforces showing every intermediate step, checking dimensional consistency, and verifying answers against limiting cases.
- Five-step protocol from problem classification to verification
- Covers pure and applied math plus physics reasoning across domains
- References MATH (12,500 problems) and PhysReason benchmarks
Scientific Reasoning by the numbers
- 17 all-time installs (skills.sh)
- Ranked #1,286 of 2,065 Data Science & ML skills by installs in the Skillselion catalog
- Data as of Aug 2, 2026 (Skillselion catalog sync)
scientific-reasoning capabilities & compatibility
- Capabilities
- research
- Use cases
- research
- Pricing
- Free
What scientific-reasoning says it does
Solve mathematical and scientific reasoning problems with rigorous step-by-step proofs and physical intuition.
Show all intermediate steps; never skip derivations
Verify dimensional consistency in physics problems
npx skills add https://github.com/beita6969/scienceclaw --skill scientific-reasoningAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Installs | 17 |
|---|---|
| repo stars | ★ 869 |
| Last updated | June 8, 2026 |
| Repository | beita6969/scienceclaw ↗ |
What it does
Solve a math or physics problem with rigorous step-by-step derivations and verification of the result.
Who is it for?
Solving competition math and physics problems that require explicit derivations.
Skip if: Retrieving documents or generating manuscripts.
When should I use this skill?
Solving pure or applied math, physics reasoning, or cross-domain mathematical modeling problems.
What you get
A step-by-step solution with justified derivations, dimensional checks, and verification against limiting cases.
- step-by-step solution
- formal proof
- verification against limiting cases
By the numbers
- five-step reasoning protocol
- MATH dataset of 12,500 problems referenced
Files
Scientific Reasoning & Problem Solving
Purpose
Solve mathematical and scientific reasoning problems with rigorous step-by-step proofs and physical intuition.
Key Datasets
- MATH Dataset (hendrycks/competition_math): 12,500 competition math problems across 7 categories (Algebra, Counting & Probability, Geometry, Intermediate Algebra, Number Theory, Prealgebra, Precalculus), difficulty levels 1-5
- PhysReason (zhibei1204/PhysReason): Physics reasoning benchmark requiring multi-step physical reasoning
Protocol
1. Problem classification — Identify domain (algebra, calculus, geometry, mechanics, thermodynamics, etc.) 2. Known-unknowns analysis — List given quantities, target quantity, constraints 3. Strategy selection — Choose approach (direct computation, proof by contradiction, dimensional analysis, symmetry arguments) 4. Step-by-step solution — Show each derivation step with justification 5. Verification — Check units, boundary conditions, limiting cases
Problem Categories
- Pure mathematics: Number theory, combinatorics, algebra, analysis, geometry
- Applied mathematics: Differential equations, optimization, numerical methods
- Physics reasoning: Mechanics, E&M, thermodynamics, quantum, relativity
- Cross-domain: Mathematical modeling of physical systems
Rules
- Show all intermediate steps; never skip derivations
- Verify dimensional consistency in physics problems
- Check answer against limiting cases and physical intuition
- For competition problems, aim for elegant solutions, not just correct ones
- Distinguish between exact and approximate answers
Related skills
Microsoft FoundryDeploy, evaluate, and continuously improve Microsoft Foundry agents from a single agent interface.490k1.3k
Ai Research ReproductionOrchestrate trustworthy, auditable reproduction of deep learning repositories directly from their READMEs.176k511Run TrainSafely execute selected deep learning training commands with standardized evidence capture.176k511Explore RunSafely run isolated exploratory experiments with clear recording and conservative selection before committing changes.176k511Paper Context ResolverFetch precise reproduction-critical details like dataset splits, preprocessing steps, or evaluation protocols from the original academic paper when the repo README leav141k511Repo Intake And PlanScan unfamiliar AI research repositories and receive a minimal, trustworthy reproduction target before investing significant time.140k511FAQ
Does it show intermediate steps?
Yes. It shows all intermediate steps and never skips derivations, with justification at each step.
How does it verify answers?
By checking units, boundary conditions, and limiting cases against physical intuition.