genpark-wukong-enterprise-brain

GenPark Wukong Enterprise Brain is a zero-dependency knowledge graph traversal, multi-hop semantic relational deduction, and data consensus reconciliation engine for autonomous enterprise agents.

flowchart TD
    A[Knowledge Graph RDF Triples] --> B[BFS Transitive Closure Solver]
    B --> C[Shortest Semantic Deduction Path]
    D[Root Entity & Predicate Sequence] --> E[Multi-Hop Relational Join Engine]
    E --> F[Resolved Terminal Entities]
    G[Contradictory Records Stream] --> H[Log-Odds Reliability Consensus]
    H --> I[True Reconciled Fact]

Features

• Knowledge Graph Pathfinding: Discovers multi-hop transitive relation chains $(s, p, o)$ using Breadth-First Search without external graph databases.
• Multi-Hop Semantic Query: Evaluates sequential predicate pipelines across structured entity graphs.
• Log-Odds Consensus Arbiter: Resolves contradictory enterprise records by weighting sources according to statistical reliability scores.

Quickstart

from client import GenParkWukongEnterpriseBrain

brain = GenParkWukongEnterpriseBrain()
kg = [{"subject": "Alice", "predicate": "leads", "object": "Team"}]
res = brain.find_knowledge_graph_relation_path(kg, "Alice", "Team")
print("Connected:", res["connected"])

License

MIT