LLM Scraper by Nova (CIVAI)

{
  "type": "object",
  "required": [
    "instruction",
    "url"
  ],
  "properties": {
    "url": {
      "type": "string",
      "description": "The starting URL for the scraping process"
    },
    "max_depth": {
      "type": "integer",
      "default": 2,
      "maximum": 3,
      "minimum": 1,
      "description": "Maximum depth of link following (1-3)"
    },
    "max_links": {
      "type": "integer",
      "default": 5,
      "maximum": 10,
      "minimum": 1,
      "description": "Maximum number of links to follow per page (1-10)"
    },
    "max_pages": {
      "type": "integer",
      "default": 10,
      "maximum": 20,
      "minimum": 1,
      "description": "Maximum total number of pages to visit (1-20)"
    },
    "instruction": {
      "type": "string",
      "description": "A natural language instruction specifying what information to find"
    },
    "js_handling": {
      "enum": [
        "minimal",
        "moderate",
        "aggressive"
      ],
      "type": "string",
      "default": "moderate",
      "description": "How to handle JavaScript-heavy pages: minimal (fastest), moderate (balanced), aggressive (most thorough)"
    },
    "timeout_seconds": {
      "type": "integer",
      "default": 120,
      "maximum": 300,
      "minimum": 30,
      "description": "Maximum duration for the entire scraping process (30-300)"
    }
  }
}