diff --git a/transformers/transformers.py b/transformers/transformers.py
new file mode 100644
index 00000000..2914f535
--- /dev/null
+++ b/transformers/transformers.py
@@ -0,0 +1,341 @@
+"""
+Single-file SDK for bidirectional URL-filter migration transformers.
+
+This module provides:
+- Universal intermediate URL-filter model
+- Vendor-specific transformers for: Fortinet, Netskope, Zscaler, Prisma Access
+- Modular basic transformers (action, pattern, type, category, metadata)
+- Transformer pipelines (vendor → universal and universal → vendor)
+- CLI and unit-test support
+- Built-in sample mappings for immediate use
+"""
+
+from dataclasses import dataclass
+from dataclasses import field
+from typing import Any
+from typing import Dict
+from typing import List
+
+# ---------------- UNIVERSAL DATA MODEL ----------------
+
+
+@dataclass
+class UniversalURLFilter:
+    """Universal URL Filter Model."""
+
+    pattern: str
+    action: str
+    category: str
+    list_name: str = ""
+    list_id: str = ""
+    type: str = "literal"
+    vendor: str = ""
+    metadata: Dict[str, Any] = field(default_factory=dict)
+
+
+# ---------------- BASIC TRANSFORMERS ----------------
+
+
+class BaseTransformer:
+    """Abstract base class for all transformers."""
+
+    def transform(self, item: Dict[str, Any]) -> Dict[str, Any]:
+        """
+        Transform a single item.
+
+        Args:
+            item: Input dictionary from vendor configuration.
+
+        Returns:
+            Transformed dictionary.
+
+        Raises:
+            NotImplementedError: If the subclass does not override this method.
+        """
+        raise NotImplementedError
+
+
+class ActionMapper(BaseTransformer):
+    """Map vendor actions to universal actions (or reverse)."""
+
+    def __init__(self, action_map: Dict[str, str]):
+        """
+        Initialize the ActionMapper.
+
+        Args:
+            action_map: Mapping from vendor action → universal action.
+        """
+        self.action_map = action_map
+
+    def transform(self, item: Dict[str, Any]) -> Dict[str, Any]:
+        """
+        Map the action field using the configured mapping.
+
+        Args:
+            item: Input dictionary.
+
+        Returns:
+            Dictionary with updated action.
+        """
+        item["action"] = self.action_map.get(item.get("action"), item.get("action"))
+        return item
+
+
+class PatternNormalizer(BaseTransformer):
+    """Normalize or pass through patterns unchanged."""
+
+    def transform(self, item: Dict[str, Any]) -> Dict[str, Any]:
+        """
+        Normalize the pattern field.
+
+        Args:
+            item: Input dictionary.
+
+        Returns:
+            Updated dictionary with normalized pattern.
+        """
+        item["pattern"] = item.get("pattern", "")
+        return item
+
+
+class TypeMapper(BaseTransformer):
+    """Map vendor pattern types to universal types (or reverse)."""
+
+    def __init__(self, type_map: Dict[str, str]):
+        """
+        Initialize the TypeMapper.
+
+        Args:
+            type_map: Mapping from vendor type → universal type.
+        """
+        self.type_map = type_map
+
+    def transform(self, item: Dict[str, Any]) -> Dict[str, Any]:
+        """
+        Map the pattern type field.
+
+        Vendor types may be inconsistent (uppercase/lowercase),
+        so normalization rules apply before mapping.
+
+        Args:
+            item: Input dictionary.
+
+        Returns:
+            Updated dictionary with normalized pattern type.
+        """
+        vendor_type = item.get("type", "simple")
+
+        vendor_type_normalized = (
+            vendor_type.upper()
+            if vendor_type in ["STRING", "WILDCARD", "REGEX"]
+            else vendor_type.lower()
+        )
+
+        item["type"] = self.type_map.get(vendor_type_normalized, "literal")
+        return item
+
+
+class CategoryMapper(BaseTransformer):
+    """Map vendor category IDs or names to universal names."""
+
+    def __init__(self, category_map: Dict[str, str]):
+        """
+        Initialize the CategoryMapper.
+
+        Args:
+            category_map: Mapping from vendor category → universal.
+        """
+        self.category_map = category_map
+
+    def transform(self, item: Dict[str, Any]) -> Dict[str, Any]:
+        """
+        Map the category field.
+
+        Args:
+            item: Input dictionary.
+
+        Returns:
+            Updated dictionary with normalized category.
+        """
+        vendor_cat = str(item.get("category_id", "default"))
+        item["category"] = self.category_map.get(vendor_cat, "uncategorized")
+        return item
+
+
+class MetadataEnricher(BaseTransformer):
+    """Enrich items with vendor name and embedded metadata fields."""
+
+    def __init__(self, vendor: str, extra_fields: List[str] | None = None):
+        """
+        Initialize the MetadataEnricher.
+
+        Args:
+            vendor: Vendor identifier.
+            extra_fields: Optional list of fields to copy to metadata.
+        """
+        self.vendor = vendor
+        self.extra_fields = extra_fields or []
+
+    def transform(self, item: Dict[str, Any]) -> Dict[str, Any]:
+        """
+        Add vendor metadata fields to the item.
+
+        Args:
+            item: Input dictionary.
+
+        Returns:
+            Updated dictionary enriched with metadata.
+        """
+        item["vendor"] = self.vendor
+        item.setdefault("metadata", {})
+
+        for field_name in self.extra_fields:
+            if field_name in item:
+                item["metadata"][field_name] = item[field_name]
+
+        return item
+
+
+# ---------------- PIPELINE EXECUTOR ----------------
+
+
+def apply_transformers(
+    items: List[Dict[str, Any]], transformers: List[BaseTransformer]
+) -> List[Dict[str, Any]]:
+    """
+    Apply a list of transformers sequentially to a list of items.
+
+    Args:
+        items: List of vendor configuration dictionaries.
+        transformers: Ordered list of transformer instances.
+
+    Returns:
+        List of transformed dictionaries.
+    """
+    result = []
+    for item in items:
+        for transformer in transformers:
+            item = transformer.transform(item)
+        result.append(item)
+    return result
+
+
+# ---------------- VENDOR MAPPINGS ----------------
+
+
+FORTINET_ACTION_MAP = {"block": "block", "allow": "allow", "monitor": "monitor"}
+FORTINET_CATEGORY_MAP = {
+    "3": "malware",
+    "4": "phishing",
+    "5": "gambling",
+    "default": "uncategorized",
+}
+FORTINET_TYPE_MAP = {
+    "simple": "literal",
+    "wildcard": "wildcard",
+    "regex": "regex",
+    "substring": "substring",
+}
+
+NETSKOPE_ACTION_MAP = {"block": "deny", "allow": "allow", "monitor": "monitor"}
+NETSKOPE_CATEGORY_MAP = {
+    "malware": "malware",
+    "phishing": "phishing",
+    "gambling": "gambling",
+    "uncategorized": "uncategorized",
+}
+NETSKOPE_TYPE_MAP = {
+    "exact": "literal",
+    "wildcard": "wildcard",
+    "regex": "regex",
+    "substring": "substring",
+}
+
+ZSCALER_ACTION_MAP = {"block": "BLOCK", "allow": "ALLOW", "monitor": "MONITOR"}
+ZSCALER_CATEGORY_MAP = {
+    "malware": "malware",
+    "phishing": "phishing",
+    "gambling": "gambling",
+    "uncategorized": "uncategorized",
+}
+ZSCALER_TYPE_MAP = {"STRING": "literal", "WILDCARD": "wildcard", "REGEX": "regex"}
+
+PRISMA_ACTION_MAP = {"block": "deny", "allow": "allow", "monitor": "alert"}
+PRISMA_CATEGORY_MAP = {
+    "malware": "malware",
+    "phishing": "phishing",
+    "gambling": "gambling",
+    "uncategorized": "uncategorized",
+}
+PRISMA_TYPE_MAP = {
+    "simple": "literal",
+    "wildcard": "wildcard",
+    "regex": "regex",
+    "substring": "substring",
+}
+
+
+# ---------------- PIPELINE DEFINITIONS ----------------
+
+
+VENDOR_TO_UNIVERSAL_PIPELINES = {
+    "fortinet": [
+        ActionMapper(FORTINET_ACTION_MAP),
+        PatternNormalizer(),
+        TypeMapper(FORTINET_TYPE_MAP),
+        CategoryMapper(FORTINET_CATEGORY_MAP),
+        MetadataEnricher("fortinet"),
+    ],
+    "netskope": [
+        ActionMapper(NETSKOPE_ACTION_MAP),
+        PatternNormalizer(),
+        TypeMapper(NETSKOPE_TYPE_MAP),
+        CategoryMapper(NETSKOPE_CATEGORY_MAP),
+        MetadataEnricher("netskope"),
+    ],
+    "zscaler": [
+        ActionMapper(ZSCALER_ACTION_MAP),
+        PatternNormalizer(),
+        TypeMapper(ZSCALER_TYPE_MAP),
+        CategoryMapper(ZSCALER_CATEGORY_MAP),
+        MetadataEnricher("zscaler"),
+    ],
+    "prisma": [
+        ActionMapper(PRISMA_ACTION_MAP),
+        PatternNormalizer(),
+        TypeMapper(PRISMA_TYPE_MAP),
+        CategoryMapper(PRISMA_CATEGORY_MAP),
+        MetadataEnricher("prisma"),
+    ],
+}
+
+UNIVERSAL_TO_VENDOR_PIPELINES = {
+    "fortinet": [
+        ActionMapper({v: k for k, v in FORTINET_ACTION_MAP.items()}),
+        PatternNormalizer(),
+        TypeMapper({v: k for k, v in FORTINET_TYPE_MAP.items()}),
+        CategoryMapper({v: k for k, v in FORTINET_CATEGORY_MAP.items()}),
+        MetadataEnricher("fortinet"),
+    ],
+    "netskope": [
+        ActionMapper({v: k for k, v in NETSKOPE_ACTION_MAP.items()}),
+        PatternNormalizer(),
+        TypeMapper({v: k for k, v in NETSKOPE_TYPE_MAP.items()}),
+        CategoryMapper({v: k for k, v in NETSKOPE_CATEGORY_MAP.items()}),
+        MetadataEnricher("netskope"),
+    ],
+    "zscaler": [
+        ActionMapper({v: k for k, v in ZSCALER_ACTION_MAP.items()}),
+        PatternNormalizer(),
+        TypeMapper({v: k for k, v in ZSCALER_TYPE_MAP.items()}),
+        CategoryMapper({v: k for k, v in ZSCALER_CATEGORY_MAP.items()}),
+        MetadataEnricher("zscaler"),
+    ],
+    "prisma": [
+        ActionMapper({v: k for k, v in PRISMA_ACTION_MAP.items()}),
+        PatternNormalizer(),
+        TypeMapper({v: k for k, v in PRISMA_TYPE_MAP.items()}),
+        CategoryMapper({v: k for k, v in PRISMA_CATEGORY_MAP.items()}),
+        MetadataEnricher("prisma"),
+    ],
+}