What problem does it solve?
Exploring how entities in a knowledge graph relate to each other—dependencies, impact chains, and indirect relationships—requires manual multi-hop queries that are slow and error-prone. This Skill automates pathfinder-style traversal from a seed entity, scoring and pruning paths so you get the most relevant connections.
Core Features & Use Cases
- Seed-based traversal: Looks up a target entity and recursively expands outward through causal edges to a configurable depth (default 3).
- Relevance scoring and pruning: Computes cumulative path scores from edge weights and keyword similarity, prunes paths below 0.3, and ranks the remainder.
- Synthesized reporting: Combines the top paths into a coherent summary and reports the top 10 paths with entity chains, relation types, and scores.
- Use Case: Before refactoring a core module, traverse the knowledge graph from that module's entity to discover every dependent component and indirect relationship that could be impacted.
Quick Start
Ask the AI to traverse the knowledge graph starting from a specific entity name, optionally specifying a traversal depth, to see its ranked dependency and relationship paths.
