Configure individual nodes

Node configuration turns a generic graph component into a more specific design choice. Each node type has its own set of controls relevant to its conceptual role.

Open configuration #

Select a node in the canvas and open its configuration dialog. Review the visible fields, descriptions, and validation messages. Modify only the values you understand and return to the graph to see whether the configuration remains valid.

Configuration categories #

Processing nodes cover choices such as text splitting and representations. Retrieval nodes cover how relevant information would be selected. Output-stage nodes describe the model and answer-generation responsibilities. Exact fields depend on the node type and should be taken from the current editor, not inferred from a generic RAG tutorial.

Provider settings are previews #

A provider or model selector in the current frontend does not mean Ragify has stored a secret or established a server connection. Do not paste production credentials into the preview. No working backend provider credential exchange is documented.

Validate after changes #

After adjusting a node, use the editor's validation tool. Validation can detect supported missing or incompatible settings but cannot confirm that a third-party provider will accept the configuration at runtime.

Related reference #

See Node reference for the purpose and expected position of each node.