unicorn mode

Author: m1rl0k

README.md

@@ -249,6 +249,10 @@ scripts/ctx.py "add error handling" --unicorn --language python
 
 **How it works:**
 
+<p align="center">
+  <img src="enhance.png" alt="Unicorn Usage" width="50%"/>
+</p>
+
 Unicorn mode uses multiple LLM passes with progressively richer code context:
 
 1. **Pass 1 (Draft)**: Retrieves rich code snippets (8 lines of context per match) to understand the codebase and sharpen the intent

enhance.png

@@ -712,13 +712,7 @@ def enhance_unicorn(query: str, **filters) -> str:
     else:
         output = final
 
-    # Stream the final output
-    import sys
-    sys.stdout.write(output)
-    sys.stdout.write("\n")
-    sys.stdout.flush()
-
-    # Sanitize citations on the final output
+    # Sanitize citations on the final output and return
     allowed_paths2, _ = extract_allowed_citations(ctx2)
     return sanitize_citations(output.strip(), allowed_paths1.union(allowed_paths2))