diff --git a/ucsinfer/__main__.py b/ucsinfer/__main__.py
index fcf5139..ed71dea 100644
--- a/ucsinfer/__main__.py
+++ b/ucsinfer/__main__.py
@@ -14,7 +14,7 @@ from .util import ffmpeg_description, parse_ucs
 
 
 def recommend_text(text: str, ctx: InferenceContext):
-    return None
+    return ctx.classify_text_ranked(text)
 
 @click.group(epilog="For more information see "
              "<https://git.squad51.us/jamie/ucsinfer>")
@@ -22,16 +22,15 @@ def recommend_text(text: str, ctx: InferenceContext):
 def ucsinfer():
     """
     Tools for applying UCS categories to sounds using large-language Models 
-
     """
     pass
 
 
 @ucsinfer.command('recommend')
-@click.option('--text', type=Optional[str], default=None,
+@click.option('--text', default=None,
               help="Recommend a category for given text instead of reading "
               "from a file")
-@click.argument('files', nargs=-1)
+@click.argument('paths', nargs=-1)
 @click.option('--model', type=str, metavar="<model-name>", 
               default="paraphrase-multilingual-mpnet-base-v2",
               show_default=True, 
@@ -65,7 +64,17 @@ def recommend(text, paths, model):
                 "text":None,
                 "recommendations":None
                 })
-                    
+    
+    for rec in recommendations:
+        print("----------")
+        if 'path' in rec:
+            print(f"Path: {rec['path']}")
+
+        print(f"Text: {rec['text'] or '<None>'}")
+        if rec['recommendations']:
+            for r in rec['recommendations']:
+                cat, subcat, _ = ctx.lookup_category(r)
+                print(f"- {r}: {cat}-{subcat}")