Book — Use Cases

Deeper worked examples for GraphForge v0.5.0: embedded Rust/Arrow engine, Parquet-backed projects, openCypher, and analyst verbs (forge.rank, forge.cluster, forge.find, …).

For install and everyday workflows, start in the Guide. For validation notes behind these guides, see research.

Guide	What you build
Knowledge graph construction	MERGE-based entity graphs, provenance, dedup, forge.find
Network analysis	Notebook metrics, datasets, forge.rank / forge.cluster, pandas/NetworkX bridge
LLM-powered workflows	Extract → store → retrieve → synthesise loops
AI agent grounding	Ontology-backed tool and capability graphs
AI agent tool recall	Large tool registries with dependencies and permissions

Prefer project directories for persistence (GraphForge("my-graph/")), not legacy file-extension paths.