Knowledge Graph Constraint Generator#

Generate constraints for knowledge graph schemas to enforce data integrity.

This skill automatically creates structural, semantic, and property constraints suitable for RDF/OWL ontologies and Neo4j property graphs.

Quick Start#

Use When#

• Designing constraints for graph schemas
• Enforcing graph integrity rules
• Generating SHACL validation rules
• Creating Neo4j database constraints
• Defining property requirements
• Establishing data governance

Inputs#

• Graph schema definitions
• Entity models
• Relationship types
• Property specifications
• Cardinality requirements
• Integrity rules

Outputs#

• Unique constraints
• Required property constraints
• Relationship constraints
• Cardinality rules
• SHACL shapes
• Cypher constraint statements

Example#

Input Schema:

text

Student: student_id (unique), name (required), email
Course: course_code (unique), title (required), credits
Relationship: (Student)-[:ENROLLED_IN]->(Course)

Generated Constraints:

text

Unique:
  Student.student_id UNIQUE
  Course.course_code UNIQUE

Required:
  Student.name (min 1)
  Course.title (min 1)

Relationships:
  (Student)-[:ENROLLED_IN]->(Course) [min 1]

Constraint Types#

1. Unique Constraints#

Ensure properties uniquely identify nodes

cypher

CREATE CONSTRAINT student_id_unique ON (s:Student) REQUIRE s.student_id IS UNIQUE

2. Required Property Constraints#

Define mandatory properties

text

Student.name IS REQUIRED
Course.title IS REQUIRED

3. Relationship Constraints#

Define allowed graph connections

text

(Student)-[:ENROLLED_IN]->(Course)
(Professor)-[:TEACHES]->(Course)

4. Cardinality Constraints#

Define relationship occurrence limits

text

Student → ENROLLED_IN → Course (minimum: 1)
Professor → TEACHES → Course (maximum: 10)

5. Domain/Range Constraints (RDF/OWL)#

Define semantic relationship constraints

turtle

:writes rdfs:domain :Researcher ;
        rdfs:range :Paper .

6. Data Type Constraints#

Ensure property values match types

text

student_id: String
credits: Integer
gpa: Float

Execution Steps#

1. Analyze Schema – Extract entities, relationships, properties
2. Identify Keys – Find unique identifiers
3. Determine Required – Identify mandatory properties
4. Analyze Relationships – Extract connection patterns
5. Calculate Cardinality – Determine relationship bounds
6. Generate Constraints – Create constraint definitions
7. Format Output – Output as SHACL, Cypher, or RDF

Output Formats#

Neo4j Cypher#

cypher

CREATE CONSTRAINT student_id_unique ON (s:Student) REQUIRE s.student_id IS UNIQUE
CREATE INDEX ON (s:Student)(name)

SHACL Shapes#

turtle

:StudentShape a sh:NodeShape ;
  sh:targetClass :Student ;
  sh:property [
    sh:path :student_id ;
    sh:minCount 1 ;
    sh:maxCount 1
  ] .

RDF/OWL#

turtle

:writes rdfs:domain :Researcher ;
        rdfs:range :Paper ;
        rdf:type owl:ObjectProperty .

Recommended Libraries#

• RDF/OWL: rdflib, pyshacl, owlready2
• Property Graph: neo4j, py2neo, networkx
• Data Validation: pydantic, jsonschema

Best Practices#

✓ Define unique identifiers early

✓ Avoid overly restrictive constraints

✓ Keep domain/range rules clear

✓ Validate data frequently

✓ Combine database and semantic constraints

✓ Document constraint reasoning

✓ Test constraints before production

References#

See constraint-patterns.md for constraint design patterns and example-constraints.md for domain constraint examples.

---

Version: 1.0.0