LlamaIndex Pipeline

{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "$id": "https://llamaindex.ai/schemas/llamaindex/pipeline.json",
  "title": "LlamaIndex Pipeline",
  "description": "A document ingestion pipeline (index) in LlamaCloud that processes, embeds, and indexes documents for retrieval in RAG applications.",
  "type": "object",
  "required": ["id", "name", "project_id"],
  "properties": {
    "id": {
      "type": "string",
      "description": "Unique identifier of the pipeline."
    },
    "name": {
      "type": "string",
      "description": "Human-readable name of the pipeline.",
      "minLength": 1,
      "maxLength": 255
    },
    "project_id": {
      "type": "string",
      "description": "Identifier of the project the pipeline belongs to."
    },
    "status": {
      "type": "string",
      "enum": ["active", "syncing", "error", "idle"],
      "description": "Current processing status of the pipeline."
    },
    "embedding_model": {
      "type": "string",
      "description": "Name of the embedding model used for vectorization."
    },
    "sync_interval": {
      "type": "integer",
      "minimum": 0,
      "description": "Scheduled sync frequency in seconds. Zero means manual sync only."
    },
    "data_sources": {
      "type": "array",
      "description": "Data sources connected to this pipeline for automatic ingestion.",
      "items": {
        "$ref": "#/$defs/DataSource"
      }
    },
    "data_sinks": {
      "type": "array",
      "description": "Data sinks where processed content is stored.",
      "items": {
        "$ref": "#/$defs/DataSink"
      }
    },
    "created_at": {
      "type": "string",
      "format": "date-time",
      "description": "Timestamp when the pipeline was created."
    },
    "updated_at": {
      "type": "string",
      "format": "date-time",
      "description": "Timestamp when the pipeline was last updated."
    }
  },
  "$defs": {
    "DataSource": {
      "type": "object",
      "description": "A data source that provides documents for automatic ingestion into a pipeline.",
      "required": ["id", "name", "source_type"],
      "properties": {
        "id": {
          "type": "string",
          "description": "Unique identifier of the data source."
        },
        "name": {
          "type": "string",
          "description": "Human-readable name of the data source.",
          "minLength": 1,
          "maxLength": 255
        },
        "source_type": {
          "type": "string",
          "description": "Type of the data source (e.g., document, table, api, database, web)."
        },
        "project_id": {
          "type": "string",
          "description": "Identifier of the project the data source belongs to."
        },
        "config": {
          "type": "object",
          "additionalProperties": true,
          "description": "Configuration specific to the data source type."
        },
        "created_at": {
          "type": "string",
          "format": "date-time",
          "description": "Timestamp when the data source was created."
        }
      }
    },
    "DataSink": {
      "type": "object",
      "description": "A data sink that receives processed and embedded content, such as a vector database.",
      "required": ["id", "name", "sink_type"],
      "properties": {
        "id": {
          "type": "string",
          "description": "Unique identifier of the data sink."
        },
        "name": {
          "type": "string",
          "description": "Human-readable name of the data sink.",
          "minLength": 1,
          "maxLength": 255
        },
        "sink_type": {
          "type": "string",
          "description": "Type of the data sink (e.g., vector_store)."
        },
        "config": {
          "type": "object",
          "additionalProperties": true,
          "description": "Configuration specific to the data sink type."
        }
      }
    }
  }
}