meta(index): Normalize legacy catalog categories

tools/scripts/generate_index.py

@@ -205,6 +205,34 @@ FAMILY_CATEGORY_RULES = [
     ("terraform-", "devops"),
 ]
 
+CATEGORY_ALIASES = {
+    # Legacy/specialized labels normalized to broader catalog buckets
+    "ai-agents": "ai-ml",
+    "voice-agents": "ai-ml",
+    "data-ai": "ai-ml",
+    "memory": "ai-ml",
+    "api-integration": "backend",
+    "blockchain": "backend",
+    "front-end": "web-development",
+    "frontend": "web-development",
+    "app-builder": "development",
+    "code": "development",
+    "code-quality": "development",
+    "development-and-testing": "development",
+    "framework": "development",
+    "database-processing": "database",
+    "document-processing": "productivity",
+    "spreadsheet-processing": "productivity",
+    "presentation-processing": "productivity",
+    "graphics-processing": "productivity",
+    "data": "data-science",
+    "marketing": "business",
+    "planning": "workflow",
+    "project-management": "workflow",
+    "reliability": "devops",
+    "test-automation": "testing",
+}
+
 
 def tokenize(text):
     return re.findall(r"[a-z0-9]+", text.lower())
@@ -254,6 +282,13 @@ def infer_category(skill_id, skill_name, description):
 
     return best_category
 
+
+def normalize_category(category):
+    if not isinstance(category, str):
+        return category
+    normalized = category.strip().lower()
+    return CATEGORY_ALIASES.get(normalized, normalized)
+
 def normalize_yaml_value(value):
     if isinstance(value, Mapping):
         return {key: normalize_yaml_value(val) for key, val in value.items()}
@@ -359,6 +394,7 @@ def generate_index(skills_dir, output_file):
                     skill_info["description"],
                 )
                 skill_info["category"] = inferred_category or "uncategorized"
+            skill_info["category"] = normalize_category(skill_info["category"])
             
             # Fallback for description if missing in frontmatter (legacy support)
             if not skill_info["description"]:

tools/scripts/tests/test_generate_index_categories.py

@@ -20,6 +20,11 @@ generate_index = load_module("tools/scripts/generate_index.py", "generate_index_
 
 
 class GenerateIndexCategoryTests(unittest.TestCase):
+    def test_normalize_category_maps_legacy_labels(self):
+        self.assertEqual(generate_index.normalize_category("front-end"), "web-development")
+        self.assertEqual(generate_index.normalize_category("ai-agents"), "ai-ml")
+        self.assertEqual(generate_index.normalize_category("document-processing"), "productivity")
+
     def test_infer_category_returns_none_for_weak_signal(self):
         inferred = generate_index.infer_category(
             "mystery-skill",
@@ -94,6 +99,22 @@ class GenerateIndexCategoryTests(unittest.TestCase):
             self.assertEqual(categories["nested-skill"], "bundles")
             self.assertEqual(categories["playwright-skill"], "testing")
 
+    def test_generate_index_normalizes_explicit_legacy_category(self):
+        with tempfile.TemporaryDirectory() as temp_dir:
+            base = pathlib.Path(temp_dir)
+            skills_dir = base / "skills"
+            output_file = base / "skills_index.json"
+
+            legacy_dir = skills_dir / "legacy-skill"
+            legacy_dir.mkdir(parents=True)
+            (legacy_dir / "SKILL.md").write_text(
+                "---\nname: legacy-skill\ncategory: front-end\ndescription: Example\n---\nbody\n",
+                encoding="utf-8",
+            )
+
+            skills = generate_index.generate_index(str(skills_dir), str(output_file))
+            self.assertEqual(skills[0]["category"], "web-development")
+
 
 if __name__ == "__main__":
     unittest.main()