package.json

{
  "name": "distill-rag",
  "version": "1.0.0",
  "description": "A modular pipeline for dataset extraction, long-chunk indexing, and semantic search for distillation-style RAG systems.",
  "main": "index.js",

  "scripts": {
    "test": "jest",
    "rebuild:index": "bash rebuild_distill_index.sh",
    "clean": "node distill_rag_cleanup.js"
  },

  "keywords": [
    "rag",
    "distillation",
    "elasticsearch",
    "embeddings",
    "chunking",
    "nlp",
    "pipeline",
    "dataset-extraction"
  ],

  "author": "",
  "license": "Apache-2.0",

  "repository": {
    "type": "git",
    "url": "https://github.com/YOURNAME/distill-rag"
  },

  "bugs": {
    "url": "https://github.com/YOURNAME/distill-rag/issues"
  },

  "engines": {
    "node": ">=18"
  },

  "dependencies": {
    "@elastic/elasticsearch": "^8.17.0",
    "cheerio": "^1.1.2",
    "dotenv": "^16.4.7",
    "fs-extra": "^11.3.0"
  },

  "devDependencies": {
    "@jest/globals": "^30.2.0",
    "jest": "^30.2.0",
    "supertest": "^7.1.4"
  }
}