TrustGraph is an open-source semantic agent governance platform published on GitHub with about 2,800 stars. It gives AI agents a governed, verifiable knowledge foundation by combining knowledge graphs, ontologies and retrieval into one deployable system.
TrustGraph ingests documents and data, extracts structured knowledge into RDF knowledge graphs shaped by OWL ontologies and SHACL constraints, and serves that knowledge to agents through GraphRAG-style retrieval. Because the knowledge layer is explicit and inspectable, agent answers can be traced back to governed facts rather than opaque vector similarity alone, and deterministic processing pipelines make results reproducible. The platform deploys with Docker Compose or Kubernetes on any major cloud or on-premises, and works with hosted LLM APIs or local models. It is Apache-2.0 licensed and actively developed in 2026. Sample corpora, ingestion pipelines and a browser-based workbench are included so teams can explore their knowledge graphs and test agent answers before wiring the system into production applications.
TrustGraph is completely free and open source under the Apache-2.0 license. There are no paid tiers; you deploy and operate the platform on your own infrastructure.
TrustGraph is for engineering and knowledge teams building agents that must answer from controlled, auditable enterprise knowledge rather than unstructured embeddings alone.