@devilsdev/rag-pipeline-utils 2.4.5 → 2.6.0

package/.ragrc.schema.json

@@ -1,218 +1,109 @@
 {
   "$schema": "http://json-schema.org/draft-07/schema#",
   "$id": "https://raw.githubusercontent.com/DevilsDev/rag-pipeline-utils/main/.ragrc.schema.json",
-  "title": "RAG Pipeline Configuration Schema",
-  "description": "Configuration schema for @devilsdev/rag-pipeline-utils",
+  "title": "rag-pipeline configuration (.ragrc.json)",
+  "description": "Configuration read by the rag-pipeline CLI and loadRagConfig() in @devilsdev/rag-pipeline-utils. Each pipeline entry names a built-in component, a custom module under \"plugins\", or a plugin registered on pluginRegistry.",
   "type": "object",
-  "required": ["plugins"],
+  "required": ["pipeline"],
   "properties": {
-    "plugins": {
+    "$schema": { "type": "string" },
+    "namespace": {
+      "type": "string",
+      "minLength": 1,
+      "default": "default",
+      "description": "Logical name for this configuration."
+    },
+    "metadata": {
       "type": "object",
-      "description": "Plugin specifications for RAG pipeline components",
-      "required": ["loader", "embedder", "retriever", "llm"],
+      "description": "Free-form information about the configuration.",
       "properties": {
-        "loader": {
-          "type": "object",
-          "description": "Document loader plugins",
-          "minProperties": 1,
-          "additionalProperties": {
-            "oneOf": [
-              { "type": "string" },
-              {
-                "type": "object",
-                "required": ["name"],
-                "properties": {
-                  "name": { "type": "string" },
-                  "version": { "type": "string" },
-                  "source": {
-                    "type": "string",
-                    "enum": ["registry", "local", "git", "npm"]
-                  },
-                  "path": { "type": "string" },
-                  "config": { "type": "object" },
-                  "enabled": { "type": "boolean" }
-                }
-              }
-            ]
-          }
-        },
-        "embedder": {
-          "type": "object",
-          "description": "Text embedding plugins",
-          "minProperties": 1,
-          "additionalProperties": {
-            "oneOf": [
-              { "type": "string" },
-              {
-                "type": "object",
-                "required": ["name"],
-                "properties": {
-                  "name": { "type": "string" },
-                  "version": { "type": "string" },
-                  "source": {
-                    "type": "string",
-                    "enum": ["registry", "local", "git", "npm"]
-                  },
-                  "path": { "type": "string" },
-                  "config": { "type": "object" },
-                  "enabled": { "type": "boolean" }
-                }
-              }
-            ]
-          }
-        },
-        "retriever": {
-          "type": "object",
-          "description": "Vector retrieval plugins",
-          "minProperties": 1,
-          "additionalProperties": {
-            "oneOf": [
-              { "type": "string" },
-              {
-                "type": "object",
-                "required": ["name"],
-                "properties": {
-                  "name": { "type": "string" },
-                  "version": { "type": "string" },
-                  "source": {
-                    "type": "string",
-                    "enum": ["registry", "local", "git", "npm"]
-                  },
-                  "path": { "type": "string" },
-                  "config": { "type": "object" },
-                  "enabled": { "type": "boolean" }
-                }
-              }
-            ]
-          }
-        },
-        "llm": {
-          "type": "object",
-          "description": "Language model plugins",
-          "minProperties": 1,
-          "additionalProperties": {
-            "oneOf": [
-              { "type": "string" },
-              {
-                "type": "object",
-                "required": ["name"],
-                "properties": {
-                  "name": { "type": "string" },
-                  "version": { "type": "string" },
-                  "source": {
-                    "type": "string",
-                    "enum": ["registry", "local", "git", "npm"]
-                  },
-                  "path": { "type": "string" },
-                  "config": { "type": "object" },
-                  "enabled": { "type": "boolean" }
-                }
-              }
-            ]
-          }
-        },
-        "reranker": {
-          "type": "object",
-          "description": "Result reranking plugins (optional)",
-          "minProperties": 1,
-          "additionalProperties": {
-            "oneOf": [
-              { "type": "string" },
-              {
-                "type": "object",
-                "required": ["name"],
-                "properties": {
-                  "name": { "type": "string" },
-                  "version": { "type": "string" },
-                  "source": {
-                    "type": "string",
-                    "enum": ["registry", "local", "git", "npm"]
-                  },
-                  "path": { "type": "string" },
-                  "config": { "type": "object" },
-                  "enabled": { "type": "boolean" }
-                }
-              }
-            ]
-          }
-        }
-      },
-      "additionalProperties": false
+        "name": { "type": "string" },
+        "version": { "type": "string" },
+        "description": { "type": "string" },
+        "createdAt": { "type": "string" }
+      }
     },
     "pipeline": {
+      "type": "array",
+      "description": "Components in pipeline order. At least an embedder and a retriever are needed for ingest and query.",
+      "items": { "$ref": "#/definitions/pipelineEntry" }
+    },
+    "plugins": {
       "type": "object",
-      "description": "Pipeline execution configuration",
-      "properties": {
-        "stages": {
-          "type": "array",
-          "description": "Pipeline execution stages",
-          "items": {
-            "type": "string",
-            "enum": ["loader", "embedder", "retriever", "llm", "reranker"]
-          }
-        },
-        "mode": {
+      "description": "Custom plugin modules by stage and name. Paths are relative to the configuration file; the module exports a class (constructed with the entry's config) or an object.",
+      "propertyNames": { "$ref": "#/definitions/stage" },
+      "additionalProperties": {
+        "type": "object",
+        "additionalProperties": {
           "type": "string",
-          "description": "Pipeline execution mode",
-          "enum": ["sequential", "parallel", "dag"],
-          "default": "sequential"
+          "description": "Path to the plugin module, e.g. \"./plugins/my-llm.js\"."
         }
-      }
+      },
+      "examples": [{ "llm": { "my-llm": "./plugins/my-llm.js" } }]
     },
     "performance": {
       "type": "object",
-      "description": "Performance and optimization settings",
-      "properties": {
-        "parallel": {
-          "type": "object",
-          "properties": {
-            "enabled": { "type": "boolean", "default": false },
-            "maxConcurrency": { "type": "number", "default": 5 },
-            "batchSize": { "type": "number", "default": 10 }
-          }
-        },
-        "streaming": {
-          "type": "object",
-          "properties": {
-            "enabled": { "type": "boolean", "default": false },
-            "maxMemoryMB": { "type": "number", "default": 512 },
-            "bufferSize": { "type": "number", "default": 100 }
-          }
-        }
-      }
+      "description": "Optional performance settings, preserved for plugins and tooling."
     },
     "observability": {
       "type": "object",
-      "description": "Logging and monitoring configuration",
+      "description": "Optional logging and metrics settings, preserved for plugins and tooling."
+    },
+    "cache": { "type": "object" },
+    "limits": { "type": "object" },
+    "storage": { "type": "object" }
+  },
+  "definitions": {
+    "stage": {
+      "type": "string",
+      "enum": ["loader", "embedder", "retriever", "reranker", "llm"]
+    },
+    "pipelineEntry": {
+      "type": "object",
+      "required": ["stage", "name"],
       "properties": {
-        "logging": {
-          "type": "object",
-          "properties": {
-            "level": {
-              "type": "string",
-              "enum": ["debug", "info", "warn", "error"],
-              "default": "info"
-            },
-            "structured": { "type": "boolean", "default": true }
-          }
+        "stage": { "$ref": "#/definitions/stage" },
+        "name": {
+          "type": "string",
+          "minLength": 1,
+          "description": "Built-in names: loader \"auto\"; embedder \"local\", \"openai\", \"ollama\", \"cohere\"; retriever \"file-store\", \"memory\"; llm \"openai\", \"anthropic\", \"ollama\"; reranker \"scoring\". Any other name must be declared under \"plugins\" or registered on pluginRegistry.",
+          "examples": [
+            "auto",
+            "local",
+            "openai",
+            "ollama",
+            "cohere",
+            "file-store",
+            "memory",
+            "anthropic",
+            "scoring"
+          ]
         },
-        "metrics": {
+        "version": { "type": "string" },
+        "config": {
           "type": "object",
-          "properties": {
-            "enabled": { "type": "boolean", "default": false },
-            "exportFormat": {
-              "type": "string",
-              "enum": ["prometheus", "json", "statsd"],
-              "default": "json"
-            }
-          }
+          "description": "Options passed to the component's constructor, e.g. { \"model\": \"gpt-4o-mini\" } or { \"path\": \".rag/index.json\" } for file-store. Do not put API keys here; use environment variables."
         }
       }
-    },
-    "namespace": {
-      "type": "string",
-      "description": "Configuration namespace for isolation"
     }
   },
-  "additionalProperties": false
+  "examples": [
+    {
+      "namespace": "default",
+      "pipeline": [
+        { "stage": "loader", "name": "auto" },
+        { "stage": "embedder", "name": "local" },
+        {
+          "stage": "retriever",
+          "name": "file-store",
+          "config": { "path": ".rag/index.json" }
+        },
+        {
+          "stage": "llm",
+          "name": "openai",
+          "config": { "model": "gpt-4o-mini" }
+        }
+      ]
+    }
+  ]
 }