Update app.py

app.py

@@ -199,29 +199,38 @@ processor = st.session_state.processor
 manager = st.session_state.manager
 
 st.title("Chunk Webpage Content Editor")
-st.caption("A tool to fetch, chunk, and refine web content.")
+st.caption("A tool to fetch, chunk, and refine web content for AI synthesis.")
 st.markdown(
     "Developed by [Emilija Gjorgjevska](https://www.linkedin.com/in/emilijagjorgjevska/). "
     "Inspired by Andrea Volpini's [work on content chunking](https://www.linkedin.com/pulse/understanding-chunking-google-ai-mode-practical-content-volpini-zseaf/)")
 
-with st.expander("ℹ️ App Information & Chunking Details", expanded=False):
+# --- MODIFIED: Added concise guidelines to the expander ---
+with st.expander("ℹ️ App Information & AI Writing Guidelines", expanded=False):
     st.info(
         """
-        • **App version:** v0.0 (alpha) — this is the very first public release, so you may run into bugs or incomplete features.
-        • **Server policy warning:** this app relies on automated requests (“bots”) under the hood.
-          If the target server enforces a restrictive bot policy (e.g., rate-limits requests, blocks unknown user-agents or IP addresses), parts of the app **may not work** as expected.
-        **What to do if you hit an issue:**
-        1. Check the server’s logs or policy settings to ensure it allows automated clients.
-        2. Keep an eye out for updates — v0.x → v1.0 is coming soon!
+        ### How Layout-Based Chunking is Implemented Here
+        This app uses a two-step process to create meaningful chunks based on a document’s structure:
+        1.  **Structural Preservation (HTML → Markdown):** It converts a webpage’s HTML into Markdown, preserving the original layout and hierarchy (e.g., `<h1>` becomes `#`).
+        2.  **Layout-Aware Parsing (`MarkdownNodeParser`):** It then uses LlamaIndex’s `MarkdownNodeParser` to split the Markdown at logical boundaries (like headers), yielding context-aware chunks that respect the original sections.
+        
         ---
-        **How Layout-Based Chunking is Implemented Here**
-          This app uses a sophisticated, two-step process to create meaningful chunks based on the document’s visual and semantic structure:
-          1. **Structural Preservation (HTML → Markdown):**
-             Converts the webpage’s HTML into Markdown, translating tags (`<h1>`, `<p>`, `<ul>`) into their Markdown equivalents (`#`, paragraph breaks, `*`) to preserve layout and hierarchy.
-          2. **Layout-Aware Parsing (`MarkdownNodeParser`):**
-             Uses LlamaIndex’s `MarkdownNodeParser` to read the structured Markdown and split it at logical boundaries (headers like `#`, `##`, etc.), yielding context-aware chunks that respect original sections.
-          """
-    , icon="ℹ️")
+
+        ### Writing for AI Verifiability: A Quick Guide
+        To ensure your content is selected and cited by AI, focus on making each chunk clear, coherent, and verifiable.
+
+        * **Structure with Headers:** Use a logical hierarchy of headings (H1, H2, H3) in your source content. The app uses these to create chunks.
+        * **Write for Clarity:**
+            * Use short, direct sentences.
+            * State facts explicitly—don't make the AI guess.
+            * Follow the "one idea per paragraph" rule to create self-contained, meaningful chunks.
+        * **Create Verifiable Blocks:** Format content as direct definitions, Q&A sections, or step-by-step guides. These are ideal formats for AI to extract and use as answers.
+        * **Use the Editor's Metrics:** In the "Chunk Editor" tab, use the real-time stats to guide your writing.
+            * **Reading Ease:** Aim for a score **above 60**.
+            * **Word Count:** Keep chunks within the target range (e.g., 40-600 words).
+            * The colors (red/green) will show if you are meeting the targets set in the "Settings" tab.
+        """
+    , icon="💡")
+
 
 url_input = st.text_input("Enter a webpage URL to start", key="url_input")
 
@@ -276,7 +285,7 @@ with tab1:
             selected_chunk = manager.get_chunk_by_id(st.session_state.selected_chunk_id)
 
             if selected_chunk:
-                # --- RESTORED: Side-by-side layout for editor and live preview ---
+                # --- Side-by-side layout for editor and live preview ---
                 editor_col, preview_col = st.columns(2)
 
                 with editor_col: