diff --git a/README.md b/README.md
index 7799c92..e8a22de 100644
--- a/README.md
+++ b/README.md
@@ -155,7 +155,17 @@ A utility node that passes an image through unchanged but extracts its width and
 
 ---
 
-### 7. ✖️ Multiplication (Dual & Latent)
+### 7. 📏 Image Latent Resolution Extractor
+**Get pixel dimensions from a Latent.**
+
+Standard Stable Diffusion latents are compressed (1/8th scale). This node reads the latent dimensions and multiplies them by 8 to give you the effective pixel width and height.
+
+> **Use Case:**
+> Useful when you have a latent and need to know its target pixel size for downstream resizing or conditionings, without decoding it first.
+
+---
+
+### 8. ✖️ Multiplication (Dual & Latent)
 **Mathematical scaling for custom upscaling.**
 
 Takes two integer inputs (like Width and Height) and an optional Latent, and multiplies them by a float factor.
@@ -176,7 +186,7 @@ Takes two integer inputs (like Width and Height) and an optional Latent, and mul
 
 ---
 
-### 8. 🧹 VRAM/RAM Cleaner
+### 9. 🧹 VRAM/RAM Cleaner
 **Manage your resources mid-workflow.**
 
 A "Pass-through" node that can connect to **any** input (Model, VAE, Image, etc.). It passes the data through unchanged but triggers a garbage collection and VRAM flush event when executed.
@@ -188,7 +198,7 @@ A "Pass-through" node that can connect to **any** input (Model, VAE, Image, etc.
 
 ---
 
-### 9. 🎮 Flow Control (Sidecar Bypass)
+### 10. 🎮 Flow Control (Sidecar Bypass)
 **Toggle nodes On/Off remotely.**
 
 A control panel that connects to up to 4 other nodes. This acts as a "Sidecar" — it does not sit in the flow of data, but rather controls the logic of the connected nodes.
@@ -201,7 +211,7 @@ A control panel that connects to up to 4 other nodes. This acts as a "Sidecar" 
 
 ---
 
-### 10. 🔀 Any Passthrough
+### 11. 🔀 Any Passthrough
 **Universal rerouting tools.**
 
 Use these to clean up "spaghetti" wires or create fallback logic. These nodes accept **any** connection type (Model, Image, String, Latent, etc.).
@@ -211,21 +221,34 @@ Use these to clean up "spaghetti" wires or create fallback logic. These nodes ac
 
 ---
 
-### 11. ⏱️ Add Delay
+### 12. ⏱️ Add Delay
 **Pause execution for debugging.**
 
 A simple node that connects to any data type, waits for a specified number of seconds, and then passes the data through unchanged. Useful for timing API calls or debugging complex flows.
 
 ---
 
-### 12. 📺 Show Text
+### 13. 🐞 Debug Data (Any Input)
+**Inspect your data connections.**
+
+Not sure what shape a tensor is? Want to check the exact resolution of a latent? Connect **ANY** output to this node. It analyzes the data and displays a summary directly on the node.
+
+> **Features:**
+> *   **Images:** Displays Resolution, Aspect Ratio, Batch Size, and Channels.
+> *   **Latents:** Displays Latent Shape and equivalent Pixel Resolution (8x).
+> *   **General:** Displays Type, Tensor Shape, Dtype, and Device.
+> *   **On-Graph Display:** Shows the data in a read-only "Matrix green" text box directly on the node itself.
+
+---
+
+### 14. 📺 Show Text
 **Display text directly on the graph.**
 
 Connect any string output (like the `Prompt Queue` or `Camera Angle Control`) to this node. It will display the text content inside a widget box on the node itself. Useful for verifying what is actually being sent to your CLIP encoder.
 
 ---
 
-### 13. 🎨 CLIP Text Encode (Prompt Enrichment)
+### 15. 🎨 CLIP Text Encode (Prompt Enrichment)
 **Standard conditioning with built-in style power.**
 
 Why copy-paste "masterpiece, best quality, 4k..." every time? This node acts exactly like the standard CLIP Text Encode but includes a dropdown with **20 curated style presets**.
@@ -246,7 +269,7 @@ Why copy-paste "masterpiece, best quality, 4k..." every time? This node acts exa
 
 ---
 
-### 14. 🖱️ Visual Marquee (Interactive)
+### 16. 🖱️ Visual Marquee (Interactive)
 **Pause, Select, and Refine.**
 
 A Human-in-the-Loop tool that pauses your workflow and opens a popup editor over your generated image. You can draw a crop box to select a specific area, then resume the workflow to process just that area.
@@ -269,7 +292,7 @@ A Human-in-the-Loop tool that pauses your workflow and opens a popup editor over
 
 ---
 
-### 15. 🧩 Tile Compositor (Merge)
+### 17. 🧩 Tile Compositor (Merge)
 **Seamlessly paste refined details back.**
 
 The partner node to **Visual Marquee**. After you have processed/upscaled/inpainted your cropped area, this node puts it back into the original full-size image at the exact correct coordinates.
@@ -280,7 +303,7 @@ The partner node to **Visual Marquee**. After you have processed/upscaled/inpain
 
 ---
 
-### 16. 💎 Detail Enhancer
+### 18. 💎 Detail Enhancer
 **Inject texture and crispness.**
 
 A specialized node that manipulates the noise schedule (Sigmas) to force the model to generate more high-frequency detail. It works by "slowing down" the sampling process during the critical detail-formation phase and artificially inflating noise values.
@@ -302,7 +325,7 @@ A specialized node that manipulates the noise schedule (Sigmas) to force the mod
 
 ---
 
-### 17. ☁️ LoRA Online
+### 19. ☁️ LoRA Online
 **Use LoRAs directly from URLs.**
 
 Skip the manual download process. Paste a link from Civitai, Tensor.art, or HuggingFace, and this node will download the LoRA, load it into your model, and optionally delete it after generation to save disk space.
@@ -359,6 +382,11 @@ Skip the manual download process. Paste a link from Civitai, Tensor.art, or Hugg
 4. Feed the Result, the Original Image, and the `tile_data` into **Tile Compositor**.
 5. Result: The original image with a high-res, refined detail pasted back in seamlessly.
 
+### The "Debugger"
+1. Not sure if your latent is the right size?
+2. Connect your Latent to the **Debug Data** node.
+3. It instantly shows: `Latent Size: 128x128` and `Pixel Res: 1024x1024` right on the graph.
+
 <br>
 
 ## 🤝 Contributing
diff --git a/__init__.py b/__init__.py
index ee2f56e..d0ded8e 100644
--- a/__init__.py
+++ b/__init__.py
@@ -30,6 +30,11 @@ try:
 except ImportError:
     EXTRACT_CLASSES, EXTRACT_DISPLAY = {}, {}
 
+try:
+    from .image_latent_resolution_extractor_node import NODE_CLASS_MAPPINGS as LAT_EXT_CLASSES, NODE_DISPLAY_NAME_MAPPINGS as LAT_EXT_DISPLAY
+except ImportError:
+    LAT_EXT_CLASSES, LAT_EXT_DISPLAY = {}, {}
+
 try:
     from .multiplication_node import NODE_CLASS_MAPPINGS as MULT_CLASSES, NODE_DISPLAY_NAME_MAPPINGS as MULT_DISPLAY
 except ImportError:
@@ -50,6 +55,12 @@ try:
 except ImportError:
     ANYPT_CLASSES, ANYPT_DISPLAY = {}, {}
 
+# New Debug Node
+try:
+    from .output_any_debug_data_node import NODE_CLASS_MAPPINGS as DEBUG_CLASSES, NODE_DISPLAY_NAME_MAPPINGS as DEBUG_DISPLAY
+except ImportError:
+    DEBUG_CLASSES, DEBUG_DISPLAY = {}, {}
+
 try:
     from .delay_node import NODE_CLASS_MAPPINGS as DELAY_CLASSES, NODE_DISPLAY_NAME_MAPPINGS as DELAY_DISPLAY
 except ImportError:
@@ -65,25 +76,21 @@ try:
 except ImportError:
     ENRICH_CLASSES, ENRICH_DISPLAY = {}, {}
 
-# Tile Compositor (and Manager)
 try:
     from .tile_manager_nodes import NODE_CLASS_MAPPINGS as TILE_CLASSES, NODE_DISPLAY_NAME_MAPPINGS as TILE_DISPLAY
 except ImportError:
     TILE_CLASSES, TILE_DISPLAY = {}, {}
 
-# Visual Marquee (Popup/JS based)
 try:
     from .visual_marquee_node import NODE_CLASS_MAPPINGS as VIS_CLASSES, NODE_DISPLAY_NAME_MAPPINGS as VIS_DISPLAY
 except ImportError:
     VIS_CLASSES, VIS_DISPLAY = {}, {}
 
-# Detail Enhancer
 try:
     from .detail_enhance_node import NODE_CLASS_MAPPINGS as DETAIL_CLASSES, NODE_DISPLAY_NAME_MAPPINGS as DETAIL_DISPLAY
 except ImportError:
     DETAIL_CLASSES, DETAIL_DISPLAY = {}, {}
 
-# LoRA Online Loader
 try:
     from .lora_online_node import NODE_CLASS_MAPPINGS as LORA_ONLINE_CLASSES, NODE_DISPLAY_NAME_MAPPINGS as LORA_ONLINE_DISPLAY
 except ImportError:
@@ -97,10 +104,12 @@ NODE_CLASS_MAPPINGS = {
     **RES_CLASSES,
     **FIT_CLASSES,
     **EXTRACT_CLASSES,
+    **LAT_EXT_CLASSES,
     **MULT_CLASSES,
     **CLEAN_CLASSES,
     **FLOW_CLASSES,
     **ANYPT_CLASSES,
+    **DEBUG_CLASSES,
     **DELAY_CLASSES,
     **DT_CLASSES,
     **ENRICH_CLASSES,
@@ -117,10 +126,12 @@ NODE_DISPLAY_NAME_MAPPINGS = {
     **RES_DISPLAY,
     **FIT_DISPLAY,
     **EXTRACT_DISPLAY,
+    **LAT_EXT_DISPLAY,
     **MULT_DISPLAY,
     **CLEAN_DISPLAY,
     **FLOW_DISPLAY,
     **ANYPT_DISPLAY,
+    **DEBUG_DISPLAY,
     **DELAY_DISPLAY,
     **DT_DISPLAY,
     **ENRICH_DISPLAY,
diff --git a/camera_angle_control_node.py b/camera_angle_control_node.py
index 169d4dd..6488527 100644
--- a/camera_angle_control_node.py
+++ b/camera_angle_control_node.py
@@ -225,4 +225,4 @@ NODE_CLASS_MAPPINGS = {
 
 NODE_DISPLAY_NAME_MAPPINGS = {
     "CameraAngleControl": "Camera Angle Control",
-}
+}
\ No newline at end of file
diff --git a/image_latent_resolution_extractor_node.py b/image_latent_resolution_extractor_node.py
new file mode 100644
index 0000000..afdb3ab
--- /dev/null
+++ b/image_latent_resolution_extractor_node.py
@@ -0,0 +1,58 @@
+import torch
+
+class ImageLatentResolutionExtractorNode:
+    """
+    A utility node that takes a LATENT (samples) as input and extracts its 
+    corresponding pixel resolution (Width and Height).
+    
+    Standard Stable Diffusion latents are 1/8th the size of the pixel image.
+    This node calculates: Pixel Size = Latent Size * 8.
+    
+    - Input: samples (Latent)
+    - Outputs: Latent (Passthrough), Width (Int), Height (Int)
+    """
+
+    def __init__(self):
+        pass
+
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "samples": ("LATENT",),
+            },
+        }
+
+    RETURN_TYPES = ("LATENT", "INT", "INT")
+    RETURN_NAMES = ("latent", "width", "height")
+    FUNCTION = "extract_resolution"
+    CATEGORY = "flow-assistor"
+
+    def extract_resolution(self, samples):
+        # samples is a dictionary: {'samples': Tensor[Batch, Channels, Height, Width]}
+        latent_tensor = samples["samples"]
+        
+        # Get dimensions from the tensor
+        # Shape is [Batch, Channels, H, W]
+        l_height = latent_tensor.shape[2]
+        l_width = latent_tensor.shape[3]
+        
+        # Convert to pixel dimensions (Standard SD factor is 8)
+        # We assume square pixels
+        pixel_width = int(l_width * 8)
+        pixel_height = int(l_height * 8)
+
+        return (
+            samples,      # Pass the latent through unchanged
+            pixel_width,
+            pixel_height
+        )
+
+# Node Registration
+NODE_CLASS_MAPPINGS = {
+    "ImageLatentResolutionExtractorNode": ImageLatentResolutionExtractorNode,
+}
+
+NODE_DISPLAY_NAME_MAPPINGS = {
+    "ImageLatentResolutionExtractorNode": "Image Latent Resolution Extractor",
+}
\ No newline at end of file
diff --git a/node_list.json b/node_list.json
index 8454b26..4b544f3 100644
--- a/node_list.json
+++ b/node_list.json
@@ -5,11 +5,13 @@
   "ResolutionSelectNode": "Resolution Selector (Groups)",
   "ImageResolutionFitNode": "Image Resolution Fit",
   "ImageResolutionExtractorNode": "Image Resolution Extractor",
+  "ImageLatentResolutionExtractorNode": "Image Latent Resolution Extractor",
   "MultiplicationNode": "Multiplication (Dual & Latent)",
   "VRAMRAMCleanerNode": "VRAM/RAM Cleaner",
   "BypassControl": "Flow Control (Sidecar Bypass)",
   "AnyPassthrough6to1": "Any Passthrough (6 → 1)",
   "AnyPassthrough1to6": "Any Passthrough (1 → 6)",
+  "OutputAnyDebugDataNode": "Debug Data (Any Input)",
   "AddDelay": "Add Delay",
   "DisplayText": "Show Text",
   "CLIPTextEncodePromptEnrichment": "CLIP Text Encode (Prompt Enrichment)",
diff --git a/output_any_debug_data_node.py b/output_any_debug_data_node.py
new file mode 100644
index 0000000..b07b90c
--- /dev/null
+++ b/output_any_debug_data_node.py
@@ -0,0 +1,124 @@
+import torch
+
+# Special AnyType to allow any connection
+class AnyType(str):
+    def __ne__(self, __value: object) -> bool:
+        return False
+
+any_type = AnyType("*")
+
+class OutputAnyDebugDataNode:
+    """
+    A debug node that accepts ANY input and analyzes it.
+    - If IMAGE: Shows resolution (WxH), Aspect Ratio, and Batch Size.
+    - If LATENT: Shows Latent Dimensions and equivalent Pixel Resolution (8x).
+    - If other: Shows Type and raw string representation.
+    
+    The result is displayed as text on the node itself (via JS) and output as a string.
+    """
+
+    def __init__(self):
+        pass
+
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "any_input": (any_type,),
+            },
+            "hidden": {
+                "unique_id": "UNIQUE_ID",
+                "extra_pnginfo": "EXTRA_PNGINFO",
+            }
+        }
+
+    RETURN_TYPES = ("STRING",)
+    RETURN_NAMES = ("debug_text",)
+    FUNCTION = "analyze_data"
+    CATEGORY = "flow-assistor/debug"
+    OUTPUT_NODE = True
+
+    def analyze_data(self, any_input, unique_id=None, extra_pnginfo=None):
+        text_output = "None"
+        
+        try:
+            # 1. Handle Torch Tensors (Images, Masks, Raw Latents)
+            if isinstance(any_input, torch.Tensor):
+                shape = any_input.shape
+                # Basic info
+                text_output = f"Type: Tensor\nShape: {list(shape)}\nDtype: {any_input.dtype}\nDevice: {any_input.device}\n"
+                
+                # Heuristic for IMAGE: [Batch, Height, Width, Channels]
+                if any_input.ndim == 4:
+                    # Standard Image: Channels at end (usually 3 or 4)
+                    if shape[3] <= 4 and shape[1] > 8 and shape[2] > 8:
+                        b, h, w, c = shape
+                        ar = w / h if h > 0 else 0
+                        text_output += f"\n--- IMAGE DETECTED ---\n"
+                        text_output += f"Resolution: {w}x{h}\n"
+                        text_output += f"Aspect Ratio: {ar:.2f}\n"
+                        text_output += f"Batch Size: {b}\n"
+                        text_output += f"Channels: {c}"
+                    
+                    # Raw Latent / NCHW: Channels at dim 1
+                    elif shape[1] <= 4 and shape[2] > 8 and shape[3] > 8:
+                        b, c, h, w = shape
+                        text_output += f"\n--- RAW LATENT/NCHW ---\n"
+                        text_output += f"Latent Size: {w}x{h}\n"
+                        text_output += f"Pixel Res (~8x): {w*8}x{h*8}\n"
+                        text_output += f"Batch Size: {b}"
+
+                # Heuristic for MASK: [Batch, Height, Width]
+                elif any_input.ndim == 3:
+                     b, h, w = shape
+                     text_output += f"\n--- MASK DETECTED ---\n"
+                     text_output += f"Resolution: {w}x{h}\n"
+                     text_output += f"Batch Size: {b}"
+
+            # 2. Handle Latent Dictionary (Standard ComfyUI Latent Wrapper)
+            elif isinstance(any_input, dict) and "samples" in any_input:
+                samples = any_input["samples"]
+                if isinstance(samples, torch.Tensor):
+                    # Shape: [Batch, 4, Height, Width]
+                    shape = samples.shape
+                    b, c, h, w = shape
+                    ar = w / h if h > 0 else 0
+                    
+                    text_output = f"Type: LATENT (Dict)\nShape: {list(shape)}\n"
+                    text_output += f"\n--- RESOLUTION INFO ---\n"
+                    text_output += f"Latent Size: {w}x{h}\n"
+                    text_output += f"Pixel Res (8x): {w*8}x{h*8}\n"
+                    text_output += f"Aspect Ratio: {ar:.2f}\n"
+                    text_output += f"Batch Size: {b}"
+                else:
+                    text_output = f"Type: LATENT (Dict)\nError: 'samples' is not a tensor."
+
+            # 3. Handle Lists
+            elif isinstance(any_input, list):
+                text_output = f"Type: List\nLength: {len(any_input)}\n"
+                if len(any_input) > 0:
+                    text_output += f"Sample[0]: {str(any_input[0])[:50]}..."
+
+            # 4. Handle Basic Types
+            elif isinstance(any_input, (int, float, bool, str)):
+                 text_output = f"Type: {type(any_input).__name__}\nValue: {any_input}"
+            
+            # 5. Generic Fallback
+            else:
+                raw_str = str(any_input)
+                text_output = f"Type: {type(any_input)}\n"
+                text_output += f"Value: {raw_str[:200]}..." if len(raw_str) > 0 else "Empty"
+
+        except Exception as e:
+            text_output = f"Error analyzing data: {str(e)}"
+
+        # Return: {"ui": {KEY: [LIST]}, "result": (TUPLE)}
+        return {"ui": {"text": [text_output]}, "result": (text_output,)}
+
+NODE_CLASS_MAPPINGS = {
+    "OutputAnyDebugDataNode": OutputAnyDebugDataNode,
+}
+
+NODE_DISPLAY_NAME_MAPPINGS = {
+    "OutputAnyDebugDataNode": "Debug Data (Any Input)",
+}
\ No newline at end of file
diff --git a/pyproject.toml b/pyproject.toml
index f9605da..f0235ad 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,7 +1,7 @@
 [project]
 name = "comfyui-flow-assistor"
 description = "Essential utility nodes for ComfyUI (Prompt Queue, Camera Control, Resolution Fit, etc.)"
-version = "1.0.5"
+version = "1.0.6"
 license = { file = "LICENSE.txt" }
 dependencies = []
 tags = ["utils", "prompt", "camera", "resolution", "workflow"]
diff --git a/web/extensions/flow_assistor/debug_data.js b/web/extensions/flow_assistor/debug_data.js
new file mode 100644
index 0000000..b5ab042
--- /dev/null
+++ b/web/extensions/flow_assistor/debug_data.js
@@ -0,0 +1,76 @@
+import { app } from "/scripts/app.js";
+import { ComfyWidgets } from "/scripts/widgets.js";
+
+// Renders the debug text returned by OutputAnyDebugDataNode directly on the node.
+
+app.registerExtension({
+  name: "FlowAssistor.DebugData",
+  async beforeRegisterNodeDef(nodeType, nodeData) {
+    if (nodeData.name !== "OutputAnyDebugDataNode") return;
+
+    // Helper to Create or Update the Text Widget
+    function populate(text) {
+      // 1. Try to find the existing widget to avoid flicker
+      const widgetName = "debug_text_msg";
+      let w = this.widgets?.find((w) => w.name === widgetName);
+
+      // 2. If it doesn't exist, create it
+      if (!w) {
+        // ComfyWidgets["STRING"] adds the widget to this.widgets automatically
+        const widgetResult = ComfyWidgets["STRING"](
+          this,
+          widgetName,
+          ["STRING", { multiline: true }],
+          app
+        );
+        w = widgetResult.widget;
+        
+        // Styling to look like a debug console
+        w.inputEl.readOnly = true;
+        w.inputEl.style.opacity = 0.9;
+        w.inputEl.style.backgroundColor = "#222";
+        w.inputEl.style.color = "#0f0"; // Matrix green text
+        w.inputEl.style.fontFamily = "monospace";
+        w.inputEl.style.fontSize = "11px";
+        w.inputEl.style.padding = "4px";
+      }
+
+      // 3. Update the value
+      w.value = text ?? "Waiting for data...";
+
+      // 4. Force Resize to fit content
+      requestAnimationFrame(() => {
+        const sz = this.computeSize();
+        if (sz[0] < this.size[0]) sz[0] = this.size[0];
+        if (sz[1] < this.size[1]) sz[1] = this.size[1];
+        this.onResize?.(sz);
+        app.graph.setDirtyCanvas(true, false);
+      });
+    }
+
+    // Hook: When the node executes (during prompt run)
+    const onExecuted = nodeType.prototype.onExecuted;
+    nodeType.prototype.onExecuted = function (message) {
+      onExecuted?.apply(this, arguments);
+      // "message.text" comes from the python return {"ui": {"text": ...}}
+      // ComfyUI usually sends lists for UI elements
+      const text = message?.text?.[0] ?? message?.text ?? "No Data";
+      populate.call(this, text);
+    };
+
+    // Hook: When the node loads from a workflow (restore saved state)
+    const onConfigure = nodeType.prototype.onConfigure;
+    nodeType.prototype.onConfigure = function () {
+      onConfigure?.apply(this, arguments);
+      // Restore the last seen value if it exists in the save file
+      if (this.widgets_values && this.widgets_values.length) {
+        // The read-only widget usually ends up being the last one in saved values
+        // or we just check index 0 if it's the only widget
+        const savedText = this.widgets_values[0];
+        if (typeof savedText === "string") {
+            populate.call(this, savedText);
+        }
+      }
+    };
+  },
+});
\ No newline at end of file