Skip to main content
skill CX Codex

math-model-selector — Claude Code skill for choosing a math model

The math-model-selector skill helps identify a mathematical framework when you have a problem but aren’t sure how to formalize it. Use it to choose a direction before working through the calculations.

MIT ★ 3,951 updated 2026-01-26

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add parcadei/Continuous-Claude-v3 --skill math-model-selector -a claude-code
Open on GitHub

What math-model-selector does

The skill asks what quantity matters, how it changes, whether randomness is involved, and whether anything needs to be optimized. Based on the answers, it points to methods such as ODEs for continuous change, Bayesian approaches for epistemic uncertainty, or integer programming for discrete choices.

It also asks how precise the answer needs to be, then suggests a starting point with relevant equations, a first step, a computation tool, and a related skill. A rough estimate points toward dimensional analysis; numerical answers call for simulation. It selects a framework, rather than solving the whole problem.

What it helps with

  • identify a mathematical field for a problem
  • distinguish continuous change from discrete steps
  • choose a method for a problem involving randomness
  • select an approach to optimization
  • decide between simulation and an analytical answer

Example prompts for the agent

  • Help me choose a mathematical model for how a rumor spreads through a population.
  • I need to formalize a problem with continuous change and some randomness. Which methods should I consider?
  • I want to minimize a function with discrete choices. What mathematical approach fits?

How to install math-model-selector

1Skills CLI (Vercel)

npx skills add parcadei/Continuous-Claude-v3 --skill math-model-selector -a claude-code

2Manually: put the files in place

~/.claude/skills/math-model-selector/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/math-model-selector/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about math-model-selector

How do I install math-model-selector in Claude Code?

Run npx skills add parcadei/Continuous-Claude-v3 --skill math-model-selector -a claude-code. The skill file is at ~/.claude/skills/math-model-selector/SKILL.md.

What does math-model-selector do?

It asks about the quantity, how it changes, uncertainty, optimization, and the required level of precision. It uses those answers to recommend a mathematical framework and a starting point.

What does it recommend for continuous versus discrete change?

For continuous change, it points to ODEs; for stepwise change, it considers difference equations and recurrences.

Source text

SKILL.md · MIT
---
name: math-model-selector
description: Routes problems to appropriate mathematical frameworks using expert heuristics
---

# Math Model Selector

## When to Use

Trigger on phrases like:
- "what math should I use"
- "which mathematical framework"
- "how do I model this"
- "what kind of problem is this"
- "formalize this problem"

Use when user has a problem but doesn't know which mathematical domain applies.

## Process

Guide user through decision tree using Polya-style questions:

### 1. Identify the quantity
**Ask:** "What quantity or phenomenon are you trying to understand?"
- Physics problem -> conservation laws, differential equations
- Economics -> equilibrium, optimization
- Data patterns -> statistics, regression

### 2. Characterize change
**Ask:** "What changes, and how does it change?"
- Discrete steps -> difference equations, recurrences
- Continuous rate -> ODEs
- Rate of rate matters -> 2nd order ODEs
- Spatial variation -> PDEs

### 3. Check for uncertainty
**Ask:** "Is there randomness or uncertainty involved?"
- Deterministic -> standard analysis
- Epistemic uncertainty -> Bayesian methods
- Random process -> probability theory, stochastic processes

### 4. Optimization check
**Ask:** "Are you optimizing something?"
- Convex objective -> linear/quadratic programming
- Non-convex -> gradient descent, evolutionary methods
- Discrete choices -> combinatorics, integer programming

### 5. Answer precision
**Ask:** "What level of answer do you need?"
- Rough estimate -> dimensional analysis
- Qualitative behavior -> phase portraits, stability
- Numerical answer -> simulation
- Exact closed form -> analytical methods

## Key Questions to Ask

1. What changes? (discrete vs continuous)
2. What causes the change? (rate dependencies)
3. What's random? (uncertainty type)
4. What's being optimized? (objective function)
5. How precise? (qualitative vs quantitative)

## Output Format

```
Framework Recommendation:
- Primary: [framework name]
- Why: [one-sentence justification]

Starting Point:
- Key equations: [relevant formulas]
- Initial approach: [first step]

Tools to Use:
- [specific script or computation tool]

Related Skills:
- [domain skill to activate next]
```

## Example Interaction

```
User: I want to model how a rumor spreads through a population

Claude: Let me help you find the right framework.

Q1: What quantity are we tracking?
> Number of people who heard the rumor over time.

Q2: What causes it to change?
> Spread rate depends on who knows and who doesn't.

Q3: Is there randomness?
> Assume deterministic (average behavior) for now.

Q4: Optimizing anything?
> No, just understanding dynamics.

Q5: What precision?
> Qualitative behavior, maybe predict peak rate.

Framework Recommendation:
- Primary: First-order ODE (logistic growth)
- Why: Rate depends on current state nonlinearly

Starting Point:
- dN/dt = r * N * (1 - N/K)
- N = heard count, K = population, r = spread rate

Tools:
- sympy_compute.py solve (equilibrium)
- Phase line for qualitative behavior

Related Skills:
- math/math-intuition-builder (understand the model)
- odes-pdes/first-order-odes (solve it)
```

## Co-activates

After framework selection, suggest:
- **math-intuition-builder** - to understand the framework
- **Relevant domain skill** - for actual computation
- **math-mode** - for symbolic verification

© parcadei · MIT · parcadei/Continuous-Claude-v3

Similar skills

All skills for Science and research →

From the same set161

The whole set →
scribe scribe — a Claude Code subagent for tracking progress across sessions. It logs decisions, completed steps, and file lists into markdown handoffs and ledgers. CX Codex ★ 3,951repo stars parallel-agents The parallel-agents skill lays out a background-run and file-based workflow for coordinating agents. Use it to track parallel tasks without pulling each agent’s full response into context. CX Codex +1 ★ 3,951repo stars lebesgue-measure A skill for measure theory tasks focused on Lebesgue measure. Provides decision trees and computational tools to work through outer measure construction, the Carathéodory criterion, and properties of measurable sets. CX Codex +1 ★ 3,951repo stars second-order-odes second-order-odes guides Claude Code through analytical and numerical workflows for solving second-order ordinary differential equations. CX Codex +1 ★ 3,951repo stars first-order-odes first-order-odes — a skill for classifying and solving first-order differential equations. Guides you through equation type detection, selects a solution method, and provides commands for both analytical and numerical solving. CX Codex +1 ★ 3,951repo stars operator-theory operator-theory — a skill for Claude Code with strategies for functional analysis: bounded operator verification, spectral theory, compact operators. Runs Z3 and SymPy for proofs and computations. CX Codex +1 ★ 3,951repo stars root-finding root-finding is a Claude Code skill that selects a numerical method for finding equation or system roots based on available derivatives and a known interval. CX Codex +1 ★ 3,951repo stars banach-spaces Tackles problems in functional analysis involving Banach spaces. Includes strategies for proving key theorems: completeness, Hahn-Banach, open mapping, closed graph, and uniform boundedness principle. CX Codex +1 ★ 3,951repo stars

Project sponsors

Companies that keep this analytics open