Neo4j Schema Documenter

Purpose

Automatically generate comprehensive markdown documentation from Neo4j database schemas including node types, relationships, properties, and visual diagrams. Helps developers understand graph structure and write effective Cypher queries.

When to Use

• Onboarding new developers
• After schema changes/migrations
• Creating API documentation
• Quarterly documentation updates
• Before major refactoring
• Explaining data model to stakeholders

Your Neo4j Databases

Database	MCP Tool Prefix	Purpose
Apex	mcp__neo4j-apex__*	Regulatory Obligations
PSPS	mcp__neo4j-psps__*	LGPS Scheme Rules
CMS Handbook	mcp__neo4j-cms-handbook__*	CMS Guidance
Scheme Docs	mcp__neo4j-scheme-docs__*	Scheme Documentation

Documentation Generation Workflow

Step 1: Extract Schema

# Get schema for database
mcp__neo4j-apex__get_neo4j_schema(sample_size=1000)

# Returns:
# - Node labels
# - Relationship types
# - Properties for each node/relationship
# - Indexes
# - Constraints

Step 2: Organize Schema Information

## Schema Structure Template

### Node Labels
[List all node types]

### Relationship Types
[List all relationship types]

### Properties
[Properties for each node/relationship]

### Indexes
[Performance indexes]

### Constraints
[Uniqueness constraints]

Step 3: Generate Documentation

Documentation template:

# [Database Name] Neo4j Schema Documentation

**Generated**: [Date]
**Sample Size**: [N nodes sampled]
**Purpose**: [Database purpose]

## Overview

[Brief description of what this database contains]

**Key Statistics**:
- Node Labels: [Count]
- Relationship Types: [Count]
- Total Properties: [Count]
- Indexes: [Count]
- Constraints: [Count]

## Graph Structure

### Visual Overview

```mermaid
graph LR
    Obligation["Obligation"]
    Topic["Topic"]
    Regulation["Regulation"]
    Entity["Entity"]

    Obligation -->|RELATES_TO| Topic
    Obligation -->|DEFINED_BY| Regulation
    Obligation -->|APPLIES_TO| Entity
    Topic -->|PART_OF| Topic
```

## Node Types

### Obligation

**Purpose**: Represents regulatory obligations that pension schemes must meet

**Properties**:
| Property | Type | Required | Description |
|----------|------|----------|-------------|
| id | String | Yes | Unique identifier |
| title | String | Yes | Obligation title |
| description | String | No | Detailed description |
| deadline_type | String | No | annual, quarterly, event-driven |
| penalty | String | No | Consequences of non-compliance |
| source_regulation | String | Yes | Reference to source regulation |
| priority | Integer | No | 1-5, higher = more critical |

**Example**:
```cypher
(:Obligation {
  id: "obl_001",
  title: "Submit Scheme Return",
  description: "Annual return to TPR within 7 months of scheme year end",
  deadline_type: "annual",
  penalty: "Fine up to £5,000",
  source_regulation: "Pensions Act 2004, s58",
  priority: 5
})
```

**Relationships**:
- `RELATES_TO → Topic`: What area this obligation covers
- `DEFINED_BY → Regulation`: Source regulation
- `APPLIES_TO → Entity`: Who must comply

**Query Examples**:
```cypher
// Find all obligations related to annual reporting
MATCH (o:Obligation)-[:RELATES_TO]->(t:Topic {name: 'Annual Reporting'})
RETURN o.title, o.deadline_type, o.priority
ORDER BY o.priority DESC

// Find obligations with highest priority
MATCH (o:Obligation)
WHERE o.priority = 5
RETURN o.title, o.source_regulation
```

### Topic

**Purpose**: Categorizes obligations by subject area

**Properties**:
| Property | Type | Required | Description |
|----------|------|----------|-------------|
| name | String | Yes | Topic name |
| description | String | No | Topic description |
| category | String | No | High-level category |

**Example**:
```cypher
(:Topic {
  name: "Annual Reporting",
  description: "Requirements for annual scheme reporting",
  category: "Governance"
})
```

**Relationships**:
- `← RELATES_TO - Obligation`: Obligations in this topic
- `PART_OF → Topic`: Hierarchical topics (parent topic)

### Regulation

**Purpose**: Source regulations and legislation

**Properties**:
| Property | Type | Required | Description |
|----------|------|----------|-------------|
| reference | String | Yes | Regulation reference |
| title | String | Yes | Regulation title |
| effective_date | Date | No | When regulation came into force |
| url | String | No | Link to full text |

**Example**:
```cypher
(:Regulation {
  reference: "LGPS Regs 2013, Reg 30",
  title: "Normal pension age",
  effective_date: "2014-04-01",
  url: "https://www.legislation.gov.uk/..."
})
```

### Entity

**Purpose**: Entities that obligations apply to

**Properties**:
| Property | Type | Required | Description |
|----------|------|----------|-------------|
| type | String | Yes | pension scheme, trustee, employer |
| name | String | No | Specific entity name |

**Example**:
```cypher
(:Entity {
  type: "pension scheme",
  name: null
})
```

## Relationship Types

### RELATES_TO

**Direction**: `Obligation → Topic`

**Purpose**: Links obligations to subject areas

**Properties**: None

**Example**:
```cypher
(:Obligation {title: "Submit Scheme Return"})
  -[:RELATES_TO]->
(:Topic {name: "Annual Reporting"})
```

### DEFINED_BY

**Direction**: `Obligation → Regulation`

**Purpose**: Links obligations to source regulations

**Properties**: None

**Example**:
```cypher
(:Obligation {title: "Submit Scheme Return"})
  -[:DEFINED_BY]->
(:Regulation {reference: "Pensions Act 2004, s58"})
```

### APPLIES_TO

**Direction**: `Obligation → Entity`

**Purpose**: Specifies who must comply

**Properties**: None

**Example**:
```cypher
(:Obligation {title: "Submit Scheme Return"})
  -[:APPLIES_TO]->
(:Entity {type: "pension scheme"})
```

### PART_OF

**Direction**: `Topic → Topic`

**Purpose**: Hierarchical topic structure

**Properties**: None

**Example**:
```cypher
(:Topic {name: "Annual Benefit Statements"})
  -[:PART_OF]->
(:Topic {name: "Member Communications"})
```

## Indexes

Performance indexes for common queries:

```cypher
// Index on Obligation.id for fast lookups
CREATE INDEX obligation_id_idx FOR (o:Obligation) ON (o.id);

// Index on Obligation.priority for sorted queries
CREATE INDEX obligation_priority_idx FOR (o:Obligation) ON (o.priority);

// Index on Topic.name for topic searches
CREATE INDEX topic_name_idx FOR (t:Topic) ON (t.name);

// Index on Regulation.reference for regulation lookups
CREATE INDEX regulation_ref_idx FOR (r:Regulation) ON (r.reference);
```

## Constraints

Uniqueness constraints:

```cypher
// Obligation ID must be unique
CREATE CONSTRAINT obligation_id_unique FOR (o:Obligation) REQUIRE o.id IS UNIQUE;

// Topic names must be unique
CREATE CONSTRAINT topic_name_unique FOR (t:Topic) REQUIRE t.name IS UNIQUE;

// Regulation references must be unique
CREATE CONSTRAINT regulation_ref_unique FOR (r:Regulation) REQUIRE r.reference IS UNIQUE;
```

## Common Query Patterns

### Find Obligations by Priority

```cypher
MATCH (o:Obligation)
WHERE o.priority >= 4
RETURN o.title, o.description, o.priority
ORDER BY o.priority DESC
```

### Find All Obligations for a Topic

```cypher
MATCH (o:Obligation)-[:RELATES_TO]->(t:Topic)
WHERE t.name = 'Annual Reporting'
RETURN o.title, o.deadline_type, o.source_regulation
```

### Find Regulatory Source for Obligation

```cypher
MATCH (o:Obligation {title: 'Submit Scheme Return'})-[:DEFINED_BY]->(r:Regulation)
RETURN r.reference, r.title, r.effective_date
```

### Find All Topics Under Category

```cypher
MATCH (child:Topic)-[:PART_OF*]->(parent:Topic {category: 'Governance'})
RETURN child.name, parent.name
```

### Path Between Obligation and Regulation

```cypher
MATCH path = (o:Obligation)-[*1..3]-(r:Regulation)
WHERE o.id = 'obl_001'
RETURN path
```

## Schema Evolution

### Version History

| Version | Date | Changes |
|---------|------|---------|
| 1.0 | 2024-01-15 | Initial schema |
| 1.1 | 2024-03-20 | Added `Obligation.priority` property |
| 1.2 | 2024-06-10 | Added `PART_OF` hierarchical relationship for Topics |

### Planned Changes

- [ ] Add `Deadline` node type for specific deadline instances
- [ ] Add `Person` node type for responsibility assignment
- [ ] Add `Document` node type for supporting documentation

## Best Practices

### Creating Nodes

```cypher
// Always set required properties
CREATE (o:Obligation {
  id: 'obl_' + randomUUID(),  // Generate unique ID
  title: 'Obligation Title',
  source_regulation: 'Regulation Reference',
  priority: 3
})
```

### Creating Relationships

```cypher
// Use MERGE to avoid duplicates
MATCH (o:Obligation {id: 'obl_001'})
MATCH (t:Topic {name: 'Annual Reporting'})
MERGE (o)-[:RELATES_TO]->(t)
```

### Querying

```cypher
// Always use parameterized queries
MATCH (o:Obligation {id: $obligationId})
RETURN o

// Not:
MATCH (o:Obligation {id: 'obl_001'})  // Vulnerable to injection
```

## Performance Tips

1. **Use indexes** for properties used in WHERE clauses
2. **Limit results** in development: `LIMIT 10`
3. **Profile queries** to identify slow operations: `PROFILE ...`
4. **Use parameters** instead of string concatenation
5. **Index relationship types** if querying specific types frequently

## Troubleshooting

### Slow Queries

```cypher
// Use EXPLAIN to see query plan
EXPLAIN
MATCH (o:Obligation)-[:RELATES_TO]->(t:Topic)
WHERE t.name = 'Annual Reporting'
RETURN o.title

// Or PROFILE for execution stats
PROFILE
[same query]
```

### Missing Relationships

```cypher
// Find orphaned nodes (no relationships)
MATCH (o:Obligation)
WHERE NOT (o)-[]-()
RETURN o.title, o.id
```

### Duplicate Nodes

```cypher
// Find duplicate obligation IDs
MATCH (o:Obligation)
WITH o.id AS id, collect(o) AS nodes
WHERE size(nodes) > 1
RETURN id, size(nodes) AS count
```

## References

- **Neo4j Documentation**: https://neo4j.com/docs/
- **Cypher Manual**: https://neo4j.com/docs/cypher-manual/
- **APOC Procedures**: https://neo4j.com/labs/apoc/

---

**Generated by**: Claude Sonnet 4.5 via Neo4j Schema Documenter
**Last Updated**: [Date]
**Database**: [Database name]

Step 4: Generate Relationship Diagram

Mermaid diagram generator:

# Based on schema, generate mermaid graph
# Include all node types and relationship types

Example output:
```mermaid
graph TB
    Obligation[("Obligation<br/>id, title, priority")]
    Topic[("Topic<br/>name, category")]
    Regulation[("Regulation<br/>reference, title")]
    Entity[("Entity<br/>type, name")]

    Obligation -->|RELATES_TO| Topic
    Obligation -->|DEFINED_BY| Regulation
    Obligation -->|APPLIES_TO| Entity
    Topic -->|PART_OF| Topic

    style Obligation fill:#e1f5ff
    style Topic fill:#fff4e1
    style Regulation fill:#f1e1ff
    style Entity fill:#e1ffe1

### Step 5: Write Documentation File

Write to markdown file

Write( file_path="[project]/docs/neo4j-schema-[database].md", content="[generated documentation]" )

Commit to git

Bash(command=" cd [project] && git add docs/neo4j-schema-[database].md && git commit -m 'docs(neo4j): update [database] schema documentation' ")

## Documentation for Each Database

### Apex Database Documentation

Focus on:
- Regulatory obligations structure
- TPR requirements
- Compliance tracking
- Deadline management

### PSPS Database Documentation

Focus on:
- LGPS regulations
- Member benefits
- Calculation rules
- Scheme rules

### CMS Handbook Database Documentation

Focus on:
- Trustee guidance
- Best practices
- Implementation advice
- Practical examples

### Scheme Docs Database Documentation

Focus on:
- Specific scheme documentation
- Policies
- Procedures
- Member communications

## Automation

**Script to update all database docs**:

```bash
#!/bin/bash
# update-neo4j-docs.sh

DATABASES=("apex" "psps" "cms-handbook" "scheme-docs")

for db in "${DATABASES[@]}"; do
  echo "Generating documentation for $db..."

  # Call Neo4j Schema Documenter skill
  # (via Claude Code)

  echo "✓ $db documentation updated"
done

echo "All Neo4j schema documentation updated"

Schedule monthly updates:

# Add to crontab
# First day of month at 9am
0 9 1 * * /path/to/update-neo4j-docs.sh

Integration with Other Skills

• pensions-research: Reference schema docs when building queries
• lgps-scheme-amendment-analyzer: Check schema when analyzing regulations
• github-workflow: Commit schema docs to git
• conventional-commits-uk: Use proper commit format for doc updates

Quick Commands

# Generate schema for specific database
# (via MCP tool)
mcp__neo4j-apex__get_neo4j_schema(sample_size=1000)

# Write to file
# (via Write tool)

# View current schema docs
ls -la docs/neo4j-schema-*.md

# Compare schemas
diff docs/neo4j-schema-apex.md docs/neo4j-schema-psps.md

Summary

This skill provides automated Neo4j schema documentation by:

• Extracting schema from all 4 Neo4j databases
• Generating comprehensive markdown documentation
• Creating visual relationship diagrams
• Including query examples and best practices
• Documenting indexes and constraints
• Providing troubleshooting guides

Use Sonnet model - schema documentation is complex but doesn't need Opus.

Helps developers understand graph structure and write effective Cypher queries.