How do I install the Discopy Nlp skill?

Run npx skillmds@latest add manutej/discopy-nlp in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Discopy Nlp safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Discopy Nlp?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Discopy Nlp free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Discopy Nlp?

manutej (@manutej) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Discopy Nlp by manutej — Agent Skill

Name: Discopy Nlp
Author: manutej

Discopy Nlp

DisCoPy categorical quantum NLP for string diagrams, monoidal categories, and compositional semantics in Python. Use when implementing compositional distributional semantics, building string diagram computations, working with monoidal category theory in code, creating quantum-inspired NLP models, or applying categorical semantics to natural language processing.

manutej 67ae850 2 files · 5.0 KB Updated Sep 17, 2026

File contents

DisCoPy Categorical Quantum NLP

DisCoPy provides Python implementations of string diagrams and monoidal categories for compositional NLP.

Installation

pip install discopy

Core Categorical Structures

Types and Morphisms

from discopy import Ty, Box, Id, Diagram

# Types (objects in monoidal category)
n = Ty('n')  # noun
s = Ty('s')  # sentence

# Morphisms (boxes between types)
john = Box('John', Ty(), n)       # John: I → n
sleeps = Box('sleeps', n, s)      # sleeps: n → s

Composition and Tensor

# Sequential composition (;)
john_sleeps = john >> sleeps  # John sleeps: I → s

# Parallel composition (⊗)
mary = Box('Mary', Ty(), n)
john_and_mary = john @ mary  # John ⊗ Mary: I → n ⊗ n

String Diagrams

# Create and draw diagram
diagram = john >> sleeps
diagram.draw()

# Export formats
diagram.draw(path="diagram.png")

Pregroup Grammar

from discopy.grammar.pregroup import Ty, Word, Cup, Id

# Pregroup types with adjoints
n = Ty('n')
s = Ty('s')
n_r = n.r  # right adjoint
n_l = n.l  # left adjoint

# Words with grammatical types
john = Word('John', n)
sleeps = Word('sleeps', n.r @ s)
loves = Word('loves', n.r @ s @ n.l)

# Parse "John sleeps"
sentence = john @ sleeps >> Cup(n, n.r) @ Id(s)

Functorial Semantics

import numpy as np
from discopy.tensor import Tensor, Dim

# Meaning functor: Grammar → Vect
d = Dim(2)

john_vec = Tensor(dom=Dim(1), cod=d, array=np.array([1, 0]))
sleeps_mat = Tensor(dom=d, cod=Dim(1), array=np.array([0.8, 0.2]))

# Compose to get sentence meaning
result = john_vec >> sleeps_mat

Quantum Circuit Semantics

from discopy.quantum import Circuit, Ket, Bra, H, CX

# Map words to quantum circuits
john_circuit = Ket(0)
mary_circuit = Ket(1)
entangle = CX >> (H @ Circuit.id(1))

Categorical Guarantees

DisCoPy ensures:

Monoidal Laws: Associativity and unit laws for ⊗
Functoriality: Semantic maps preserve composition
String Diagram Correctness: Diagrams represent valid morphisms
Type Safety: Composition only when types match

---name: discopy-nlpdescription: DisCoPy categorical quantum NLP for string diagrams, monoidal categories, and compositional semantics in Python. Use when implementing compositional distributional semantics, building string diagram computations, working with monoidal category theory in code, creating quantum-inspired NLP models, or applying categorical semantics to natural language processing.---# DisCoPy Categorical Quantum NLPDisCoPy provides Python implementations of string diagrams and monoidal categories for compositional NLP.## Installation```bashpip install discopy```## Core Categorical Structures### Types and Morphisms```pythonfrom discopy import Ty, Box, Id, Diagram# Types (objects in monoidal category)n = Ty('n') # nouns = Ty('s') # sentence# Morphisms (boxes between types)john = Box('John', Ty(), n) # John: I → nsleeps = Box('sleeps', n, s) # sleeps: n → s```### Composition and Tensor```python# Sequential composition (;)john_sleeps = john >> sleeps # John sleeps: I → s# Parallel composition (⊗)mary = Box('Mary', Ty(), n)john_and_mary = john @ mary # John ⊗ Mary: I → n ⊗ n```## String Diagrams```python# Create and draw diagramdiagram = john >> sleepsdiagram.draw()# Export formatsdiagram.draw(path="diagram.png")```## Pregroup Grammar```pythonfrom discopy.grammar.pregroup import Ty, Word, Cup, Id# Pregroup types with adjointsn = Ty('n')s = Ty('s')n_r = n.r # right adjointn_l = n.l # left adjoint# Words with grammatical typesjohn = Word('John', n)sleeps = Word('sleeps', n.r @ s)loves = Word('loves', n.r @ s @ n.l)# Parse "John sleeps"sentence = john @ sleeps >> Cup(n, n.r) @ Id(s)```## Functorial Semantics```pythonimport numpy as npfrom discopy.tensor import Tensor, Dim# Meaning functor: Grammar → Vectd = Dim(2)john_vec = Tensor(dom=Dim(1), cod=d, array=np.array([1, 0]))sleeps_mat = Tensor(dom=d, cod=Dim(1), array=np.array([0.8, 0.2]))# Compose to get sentence meaningresult = john_vec >> sleeps_mat```## Quantum Circuit Semantics```pythonfrom discopy.quantum import Circuit, Ket, Bra, H, CX# Map words to quantum circuitsjohn_circuit = Ket(0)mary_circuit = Ket(1)entangle = CX >> (H @ Circuit.id(1))```## Categorical GuaranteesDisCoPy ensures:1. **Monoidal Laws**: Associativity and unit laws for ⊗2. **Functoriality**: Semantic maps preserve composition3. **String Diagram Correctness**: Diagrams represent valid morphisms4. **Type Safety**: Composition only when types match

Frequently asked questions