entropy — Claude Code skill: information theory entropy calculations
A skill for Shannon entropy, conditional and differential entropy, KL divergence, and maximum entropy principle problems.
MIT ★ 3,951 updated 2026-01-26
npx skills add parcadei/Continuous-Claude-v3 --skill entropy -a claude-codeWhat entropy does
The skill provides a decision tree for entropy calculations in discrete and continuous distributions. It includes ready-made scipy.stats.entropy and sympy_compute.py commands for integrals and simplifications. Allowed tools are Bash and Read.
It gives formulas for Shannon entropy H(X) = -sum p(x) log₂ p(x), joint H(X,Y), and conditional H(Y|X), along with chain rule properties and inequalities. For continuous cases, differential entropy h(X) can be negative.
The maximum entropy principle selects the least biased distribution under given constraints: uniform with no constraints, exponential for E[X] = μ, Gaussian for E[X] and Var[X].
What it helps with
- compute Shannon entropy for a discrete distribution
- find KL divergence between two distributions
- calculate differential entropy of a continuous variable
- verify conditional entropy properties H(Y|X)
- determine the maximum entropy distribution under constraints
Example prompts for the agent
- Calculate entropy for probabilities [0.25, 0.25, 0.25, 0.25]
- Find KL divergence between p = [0.5, 0.5] and q = [0.9, 0.1]
- Simplify the entropy expression
-p*log(p, 2) - (1-p)*log(1-p, 2)in SymPy
How to install entropy
1Skills CLI (Vercel)
npx skills add parcadei/Continuous-Claude-v3 --skill entropy -a claude-code2Manually: put the files in place
~/.claude/skills/entropy/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/entropy/
1Skills CLI (Vercel)
npx skills add parcadei/Continuous-Claude-v3 --skill entropy -a codex2Manually: put the files in place
.agents/skills/entropy/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add parcadei/Continuous-Claude-v3 --skill entropy -a cursor2Manually: put the files in place
.cursor/skills/entropy/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install parcadei/Continuous-Claude-v3 .claude/skills/math/information-theory/entropy2Manually: put the files in place
.github/skills/entropy/SKILL.md1Skills CLI (Vercel)
npx skills add parcadei/Continuous-Claude-v3 --skill entropy -a gemini-cli2Platform CLI
gemini skills install https://github.com/parcadei/Continuous-Claude-v3 --path .claude/skills/math/information-theory/entropyManually: put the files in place
.devin/skills/entropy/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about entropy
How to install entropy in Claude Code?
Run npx skills add parcadei/Continuous-Claude-v3 --skill entropy -a claude-code. The skill will be placed in ~/.claude/skills/entropy/SKILL.md.
What does entropy do?
A skill for information theory tasks: computes Shannon, joint, conditional, and differential entropy, KL divergence, applies the maximum entropy principle. Uses scipy.stats.entropy and sympy_compute.py.
Which tools are allowed in entropy?
Bash and Read. The skill runs Python scripts via uv run and reads documentation files.
Source text
SKILL.md · MIT
---
name: entropy
description: "Problem-solving strategies for entropy in information theory"
allowed-tools: [Bash, Read]
---
# Entropy
## When to Use
Use this skill when working on entropy problems in information theory.
## Decision Tree
1. **Shannon Entropy**
- H(X) = -sum p(x) log2 p(x)
- Maximum for uniform distribution: H_max = log2(n)
- Minimum = 0 for deterministic (one outcome certain)
- `scipy.stats.entropy(p, base=2)` for discrete
2. **Entropy Properties**
- Non-negative: H(X) >= 0
- Concave in p
- Chain rule: H(X,Y) = H(X) + H(Y|X)
- `z3_solve.py prove "entropy_nonnegative"`
3. **Joint and Conditional Entropy**
- H(X,Y) = -sum sum p(x,y) log2 p(x,y)
- H(Y|X) = H(X,Y) - H(X)
- H(Y|X) <= H(Y) with equality iff independent
4. **Differential Entropy (Continuous)**
- h(X) = -integral f(x) log f(x) dx
- Can be negative!
- Gaussian: h(X) = 0.5 * log2(2*pi*e*sigma^2)
- `sympy_compute.py integrate "-f(x)*log(f(x))" --var x`
5. **Maximum Entropy Principle**
- Given constraints, max entropy distribution is least biased
- Uniform for no constraints
- Exponential for E[X] = mu constraint
- Gaussian for E[X], Var[X] constraints
## Tool Commands
### Scipy_Entropy
```bash
uv run python -c "from scipy.stats import entropy; p = [0.25, 0.25, 0.25, 0.25]; H = entropy(p, base=2); print('Entropy:', H, 'bits')"
```
### Scipy_Kl_Div
```bash
uv run python -c "from scipy.stats import entropy; p = [0.5, 0.5]; q = [0.9, 0.1]; kl = entropy(p, q); print('KL divergence:', kl)"
```
### Sympy_Entropy
```bash
uv run python -m runtime.harness scripts/sympy_compute.py simplify "-p*log(p, 2) - (1-p)*log(1-p, 2)"
```
## Key Techniques
*From indexed textbooks:*
- [Elements of Information Theory] Elements of Information Theory -- Thomas M_ Cover & Joy A_ Thomas -- 2_, Auflage, New York, NY, 2012 -- Wiley-Interscience -- 9780470303153 -- 2fcfe3e8a16b3aeefeaf9429fcf9a513 -- Anna’s Archive. What is the channel capacity of this channel? This is the multiple\-access channel solved by Liao and Ahlswede.
## Cognitive Tools Reference
See `.claude/skills/math-mode/SKILL.md` for full tool documentation.
© parcadei · MIT · parcadei/Continuous-Claude-v3