Show HF_TOKEN detection status in Step 1

Evaluated at container startup, so a Space operator can immediately see
whether the HF_TOKEN secret reached the process or whether it's missing,
misnamed, or awaiting a restart - instead of discovering it via a failed
chat call.

app.py

@@ -320,6 +320,20 @@ with gr.Blocks(title="Document Intelligence — OCR & Structured Data Extraction
         with gr.Row(elem_id="quick-prompts"):
             quick_buttons = {label: gr.Button(label, size="sm") for label in QUICK_PROMPTS}
 
+        # Evaluated at container startup - tells the operator at a glance
+        # whether the HF_TOKEN secret actually reached this process.
+        _token_detected = bool(os.environ.get("HF_TOKEN"))
+        gr.Markdown(
+            '<div class="sample-note">Space token status: '
+            + ("<strong>HF_TOKEN detected</strong> - the assistant is ready to use."
+               if _token_detected else
+               "<strong>HF_TOKEN not detected.</strong> The secret is missing, misnamed "
+               "(it must be exactly <code>HF_TOKEN</code>, saved as a Secret, not a Variable), "
+               "or the Space hasn't been restarted since it was added. You can still paste "
+               "a token below, or skip the assistant and use the manual settings in Step 2.")
+            + "</div>"
+        )
+
         with gr.Accordion("Use your own Hugging Face token (optional)", open=False):
             gr.Markdown("Only needed if this Space's `HF_TOKEN` secret isn't set, or you want to use your own.")
             hf_token_input = gr.Textbox(label="Hugging Face token", type="password", placeholder="hf_...", show_label=False)