Update app.py

app.py

@@ -1715,15 +1715,15 @@ html, body {
   align-items: center !important;
   justify-content: center !important;
   width: auto !important;
-  min-width: 150px !important;
-  max-width: 240px !important;
-  min-height: 40px !important;
+  min-width: 170px !important;
+  max-width: 270px !important;
+  min-height: 46px !important;
   margin: 0 !important;
-  padding: 9px 20px !important;
+  padding: 10px 22px !important;
   flex: 0 0 auto !important;
   text-align: center !important;
   white-space: nowrap !important;
-  font-size: 14px !important;
+  font-size: 17px !important;
   border-radius: 9px !important;
   color: #9e93a5 !important;
 }
@@ -2043,13 +2043,15 @@ ul.options[role="listbox"] {
   text-align: center !important;
 }
 
-/* Primary tabs: give the long ML Modeling label more room and allow wrapping. */
+/* Primary tabs: give the long ML Modeling label more room and allow wrapping.
+   (Column ratios follow content, not position: 02 is now the short "ML Agent"
+   label and 03 is the long "ML Modeling (Optional Fallback Models)" label.) */
 @media (min-width: 801px) {
   #main-tabs [role="tablist"] {
     grid-template-columns:
       minmax(0, 0.88fr)
-      minmax(0, 1.24fr)
-      minmax(0, 0.88fr) !important;
+      minmax(0, 0.88fr)
+      minmax(0, 1.24fr) !important;
     gap: 14px !important;
   }
 
@@ -2272,8 +2274,29 @@ def build_app():
                                 schema_table = gr.Dataframe(label="Dataset Schema:", interactive=False, wrap=True, elem_id="dataset-schema")
                                 gr.HTML(f'''<div class="info-note"><strong>HF sample:</strong> <code>{EXAMPLE_DATASET_REPO}</code> · target <code>{EXAMPLE_DATASET_TARGET}</code>. Public Space: do not upload confidential, regulated, proprietary, production customer, or PII data.</div>''')
 
+                        # Builds the agent tab for natural-language ML tasks, generated code, and tool-trace inspection
+                        with gr.Tab("02. ML Agent", id="agent"):
+                            with gr.Column(elem_classes=["tab-body"]):
+                                gr.HTML('''<div class="section-head"><div class="section-kicker">Machine Learning Agent</div><h2>Turn Dataset Findings Into an ML Strategy and Pipeline Code (sci-kit learn)</h2><p>Agent can inspect the dataset, validate the target framing, train the selected model, compare baselines, and generate scikit-learn pipelines.</p></div>''')
+                                # Configures the agent task prompt field with a comprehensive default machine-learning request
+                                task = gr.Textbox(
+                                    label="Machine learning task",
+                                    value="Audit this dataset for supervised machine learning, confirm the target framing, train a baseline model, compare appropriate algorithms, explain the evaluation metrics, flag leakage or data risks, and generate a production-oriented scikit-learn pipeline.",
+                                    placeholder="Example: Predict customer churn, compare baseline classifiers, prioritize recall, and generate a deployable preprocessing + model pipeline.",
+                                    lines=6,
+                                    max_lines=12,
+                                )
+                                # Creates the agent execution button and result components for answers, code, and tool traces
+                                agent_btn = gr.Button("Run Machine Learning Agent", variant="primary", elem_id="agent-button")
+                                agent_answer = gr.Markdown("", elem_id="agent-answer", elem_classes=["result-markdown"])
+                                with gr.Tabs(elem_id="agent-result-tabs", elem_classes=["result-tabs"]):
+                                    with gr.Tab("Recommended Pipeline"):
+                                        agent_code = gr.Code(label="Generated Python", language="python", lines=28)
+                                    with gr.Tab("Tool Trace"):
+                                        trace = gr.JSON(label="Agent tool calls and observations")
+
                         # Builds the modeling tab for target selection, training, comparison, and pipeline generation
-                        with gr.Tab("02. ML Modeling (Optional Fallback Models)", id="model-lab"):
+                        with gr.Tab("03. ML Modeling (Optional Fallback Models)", id="model-lab"):
                             with gr.Column(elem_classes=["tab-body"]):
                                 gr.HTML('''<div class="section-head"><div class="section-kicker">Supervised Learning</div><h2>Choose a Target Variable (Feature), Algorithm, and Evaluation Setup</h2><p>Build a preprocessing + model pipeline, evaluate it on a holdout split, compare baselines, and evaluate generated pipelines (scikit-learn pipelines).</p></div>''')
 
@@ -2332,27 +2355,6 @@ def build_app():
                                 pipeline_code = gr.Code(label="Selected Algorithm Pipeline", language="python", lines=30)
                                 model_artifact = gr.File(label="Trained Pipeline Artifact (.joblib)", interactive=False)
 
-                        # Builds the agent tab for natural-language ML tasks, generated code, and tool-trace inspection
-                        with gr.Tab("03. ML Agent", id="agent"):
-                            with gr.Column(elem_classes=["tab-body"]):
-                                gr.HTML('''<div class="section-head"><div class="section-kicker">Machine Learning Agent</div><h2>Turn Dataset Findings Into an ML Strategy and Pipeline Code (sci-kit learn)</h2><p>Agent can inspect the dataset, validate the target framing, train the selected model, compare baselines, and generate scikit-learn pipelines.</p></div>''')
-                                # Configures the agent task prompt field with a comprehensive default machine-learning request
-                                task = gr.Textbox(
-                                    label="Machine learning task",
-                                    value="Audit this dataset for supervised machine learning, confirm the target framing, train a baseline model, compare appropriate algorithms, explain the evaluation metrics, flag leakage or data risks, and generate a production-oriented scikit-learn pipeline.",
-                                    placeholder="Example: Predict customer churn, compare baseline classifiers, prioritize recall, and generate a deployable preprocessing + model pipeline.",
-                                    lines=6,
-                                    max_lines=12,
-                                )
-                                # Creates the agent execution button and result components for answers, code, and tool traces
-                                agent_btn = gr.Button("Run Machine Learning Agent", variant="primary", elem_id="agent-button")
-                                agent_answer = gr.Markdown("", elem_id="agent-answer", elem_classes=["result-markdown"])
-                                with gr.Tabs(elem_id="agent-result-tabs", elem_classes=["result-tabs"]):
-                                    with gr.Tab("Recommended Pipeline"):
-                                        agent_code = gr.Code(label="Generated Python", language="python", lines=28)
-                                    with gr.Tab("Tool Trace"):
-                                        trace = gr.JSON(label="Agent tool calls and observations")
-
                 # Builds the sidebar containing live workspace status, runtime information, controls, and diagnostics
                 with gr.Column(scale=4, min_width=335, elem_classes=["side-column"]):
                     status = gr.HTML(_status(None))