Create app.py

app.py

@@ -0,0 +1,35 @@
+import streamlit as st
+import requests
+
+st.set_page_config(page_title="Qcode Agent: Qiskit Bug Fixer", layout="centered")
+st.title("🤖 Qcode Agent: LLM-Powered Quantum Code Repair")
+st.markdown("Paste your Qiskit code below and let an LLM find and fix bugs.")
+
+# Code input
+code_input = st.text_area("Qiskit Code Input", height=300, placeholder="Paste your Qiskit code here...")
+
+# Backend selection
+model_choice = st.selectbox("Choose LLM Backend", ["GPT-4 (OpenAI)", "Granite-8B-Qiskit (HF)", "Code Llama (HF)"])
+
+# API setup (mocked)
+def query_llm(model, code):
+    if model == "GPT-4 (OpenAI)":
+        return "# Fixed by GPT-4\n" + code.replace("cx(q[0], q[0])", "cx(q[0], q[1])")
+    elif model == "Granite-8B-Qiskit (HF)":
+        return "# Granite-8B fix suggestion:\n" + code.replace("cx(q[0], q[0])", "cx(q[0], q[1])")
+    elif model == "Code Llama (HF)":
+        return "# Code Llama fix suggestion:\n" + code.replace("cx(q[0], q[0])", "cx(q[0], q[1])")
+    return "No fix found."
+
+# Button to trigger repair
+if st.button("Fix My Code"):
+    if not code_input.strip():
+        st.warning("Please paste some Qiskit code to analyze.")
+    else:
+        with st.spinner("Querying the selected LLM..."):
+            fixed_code = query_llm(model_choice, code_input)
+            st.success("Patch generated successfully!")
+            st.code(fixed_code, language="python")
+
+st.markdown("---")
+st.caption("Built by your Quantum AI copilot 🧠⚛️")