codesummary/rag-schema.json

Cross-platform CLI tool that generates professional PDF documentation and RAG-optimized JSON outputs from project source code. Perfect for code reviews, audits, documentation, and AI/ML applications with semantic chunking and precision offsets.

{
  "$schema": "http://json-schema.org/draft-07/schema#",
  "$id": "https://github.com/skamoll/CodeSummary/schemas/rag-output.json",
  "title": "CodeSummary RAG Output Schema",
  "description": "Schema for CodeSummary RAG-optimized JSON output format",
  "type": "object",
  "required": ["metadata", "files", "index"],
  "properties": {
    "metadata": {
      "type": "object",
      "required": ["projectName", "generatedAt", "version", "generator"],
      "properties": {
        "projectName": { "type": "string" },
        "generatedAt": { "type": "string", "format": "date-time" },
        "version": { "type": "string", "pattern": "^\\d+\\.\\d+\\.\\d+$" },
        "generator": { "type": "string" },
        "scanPath": { "type": "string" },
        "config": { "type": "object" },
        "summary": { "type": "object" }
      }
    },
    "files": {
      "type": "array",
      "items": {
        "type": "object",
        "required": ["id", "path", "extension", "language", "hash", "chunks"],
        "properties": {
          "id": { "type": "string" },
          "path": { "type": "string" },
          "extension": { "type": "string" },
          "language": { "type": "string" },
          "size": { "type": "number", "minimum": 0 },
          "hash": { "type": "string", "pattern": "^sha256-[a-f0-9]{64}$" },
          "modified": { "type": "string", "format": "date-time" },
          "tags": { "type": "array", "items": { "type": "string" } },
          "chunks": {
            "type": "array",
            "items": {
              "type": "object",
              "required": ["id", "content", "tokenEstimate"],
              "properties": {
                "id": { "type": "string" },
                "content": { "type": "string" },
                "tokenEstimate": { "type": "number", "minimum": 0 },
                "lineStart": { "type": "number", "minimum": 1 },
                "lineEnd": { "type": "number", "minimum": 1 },
                "chunkingMethod": { "type": "string" },
                "imports": { "type": "array", "items": { "type": "string" } },
                "calls": { "type": "array", "items": { "type": "string" } }
              }
            }
          }
        }
      }
    },
    "index": {
      "type": "object",
      "required": ["version", "generatedAt", "summary", "chunkOffsets", "fileOffsets", "statistics"],
      "properties": {
        "version": { "type": "string", "pattern": "^\\d+\\.\\d+\\.\\d+$" },
        "generatedAt": { "type": "string", "format": "date-time" },
        "summary": {
          "type": "object",
          "required": ["fileCount", "chunkCount", "totalBytes"],
          "properties": {
            "fileCount": { "type": "number", "minimum": 0 },
            "chunkCount": { "type": "number", "minimum": 0 },
            "totalBytes": { "type": "number", "minimum": 0 },
            "languages": { "type": "array", "items": { "type": "string" } },
            "extensions": { "type": "array", "items": { "type": "string" } }
          }
        },
        "chunkOffsets": {
          "type": "object",
          "patternProperties": {
            "^chunk_[a-f0-9]+_\\d+$": {
              "type": "object",
              "required": ["jsonStart", "jsonEnd", "contentStart", "contentEnd", "filePath"],
              "properties": {
                "jsonStart": { "type": "number", "minimum": 0 },
                "jsonEnd": { "type": "number", "minimum": 0 },
                "contentStart": { "type": "number", "minimum": 0 },
                "contentEnd": { "type": "number", "minimum": 0 },
                "filePath": { "type": "string" }
              }
            }
          }
        },
        "fileOffsets": {
          "type": "object",
          "patternProperties": {
            "^[a-f0-9]+$": {
              "type": "array",
              "items": { "type": "number", "minimum": 0 },
              "minItems": 2,
              "maxItems": 2
            }
          }
        },
        "statistics": {
          "type": "object",
          "required": ["processingTimeMs", "bytesWritten", "totalFiles"],
          "properties": {
            "processingTimeMs": { "type": "number", "minimum": 1 },
            "bytesPerSecond": { "type": "number", "minimum": 0 },
            "bytesWritten": { "type": "number", "minimum": 0 },
            "totalFiles": { "type": "number", "minimum": 0 },
            "emptyFiles": { "type": "number", "minimum": 0 }
          }
        }
      }
    }
  }
}