lifesciences-research: Command for Claude Code

Generate comprehensive production graph statistics for Neo4j Aura analytics and capacity planning.

Instructions

Execute these queries sequentially to build a complete analytics report. Report results to the user in a formatted summary.

Query 1: Episode Distribution by Namespace

mcp__neo4j-aura-cypher__read_neo4j_cypher(
    query="""
    MATCH (e:Episodic)
    RETURN e.group_id AS namespace,
           count(*) AS episodes,
           min(e.created_at) AS first_episode,
           max(e.created_at) AS last_episode
    ORDER BY episodes DESC
    """
)

Purpose: Shows which namespaces contain data and identifies growth patterns.

Query 2: Total Entity Count

mcp__neo4j-aura-cypher__read_neo4j_cypher(
    query="""
    MATCH (n:Entity)
    RETURN count(n) AS total_entities
    """
)

Purpose: Get total entity count for summary statistics.

Query 3: Entity Type Distribution

mcp__neo4j-aura-cypher__read_neo4j_cypher(
    query="""
    MATCH (n:Entity)
    RETURN labels(n) AS entity_types,
           count(*) AS count
    ORDER BY count DESC
    LIMIT 15
    """
)

Purpose: Breakdown by entity labels (Person, Company, Concept, etc.). Helps validate extraction quality.

Query 4: Relationship Type Frequency

mcp__neo4j-aura-cypher__read_neo4j_cypher(
    query="""
    MATCH ()-[r]->()
    RETURN type(r) AS relationship_type,
           count(*) AS frequency,
           count(CASE WHEN r.invalid_at IS NOT NULL THEN 1 END) AS invalidated_count,
           round(100.0 * count(CASE WHEN r.invalid_at IS NOT NULL THEN 1 END) / count(*), 2) AS invalidation_rate_pct
    ORDER BY frequency DESC
    LIMIT 15
    """
)

Purpose: Most common relationship patterns and invalidation rates. Identifies schema usage and knowledge evolution velocity.

Query 5: Temporal Invalidation Statistics

mcp__neo4j-aura-cypher__read_neo4j_cypher(
    query="""
    MATCH ()-[r]->()
    WHERE r.invalid_at IS NOT NULL AND r.valid_at IS NOT NULL
    RETURN type(r) AS relationship_type,
           count(*) AS total_invalidated,
           round(avg(duration.between(
             datetime(r.valid_at),
             datetime(r.invalid_at)
           ).seconds) / 3600.0, 2) AS avg_lifetime_hours
    ORDER BY total_invalidated DESC
    LIMIT 10
    """
)