Configuration reference¶
OntoCast reads its settings from environment variables, or from a .env file in the working directory. This reference lists all 248 of them, generated from the settings models. To choose settings for a task, start with Configuring OntoCast.
| Page | What it controls |
|---|---|
| Pipeline and server | What a run does, how wide it fans out, and where the server listens. |
| Language model | Provider, model, credentials, request limits and the LLM response cache. |
| Document conversion | How uploaded documents become text. |
| Chunking | How text is cut into content units and which sections are kept. |
| Ontology retrieval | How each unit's slice of the ontology catalog is retrieved and sized. |
| Embeddings and vector stores | The embedding model and the LanceDB or Qdrant backend behind retrieval. |
| Facts validation | The facts critic, repairs, acceptance rules and SHACL validation. |
| Ontology validation | The ontology critic and its acceptance rules. |
| Entity disambiguation | When two extracted entities are merged into one. |
| Web search | Optional web evidence for ontology rendering and critique. |
| Storage and paths | The Fuseki triple store, seed ontologies and on-disk caches. |