new ver update doc

Author: HomunMage

Backend/index.md

@@ -2,9 +2,41 @@
 
 The core file that convert json to LangGraph is [/src/WorkFlow.py](https://github.com/LangGraph-GUI/LangGraph-GUI-backend/blob/main/src/WorkFlow.py)
 
-This use Flask server to communicate with frontend.
+[llm.py](https://github.com/LangGraph-GUI/LangGraph-GUI-backend/blob/main/src/llm.py) demo how to replace to different LLM services.
 
-Current implementation only runs graph.json, need frontend upload.
+## Json Format
+
+json node format like this:
+you can extend ext to put your data, this is elastic design.
+for example, the ext.info is a ext and able to parse in our backend.
+```
+{
+  "nodes": [
+    {
+      "uniq_id": "1",
+      "nexts": [
+        "2"
+      ],
+      "type": "START",
+      "name": "Node 1",
+      "description": "",
+      "tool": "",
+      "true_next": null,
+      "false_next": null,
+      "ext": {
+        "pos_x": 126,
+        "pos_y": 146,
+        "width": 200,
+        "height": 200,
+        "info": ""
+      }
+    },
+```
+
+
+This use FastAPI server to communicate with frontend.
+
+Current implementation only runs graph.json.
 
 *For a beginner-friendly introduction to LangGraph, visit [LangGraph-learn](https://github.com/LangGraph-GUI/LangGraph-learn).*
 

Frontend/Graph.md

@@ -40,3 +40,7 @@ CONDITION node will mapping to LangGraph conditional_edge
 ## Other Resource
 *If you want to learn more about LangGraph, we have LangGraph for dummy : [LangGraph-learn](https://github.com/LangGraph-GUI/LangGraph-learn)*
 
+#### Information Node
+INFO node do nothing, just append info to previous stage output
+
+![](images/info.webp)

Frontend/Run.md

@@ -7,16 +7,13 @@ layout: frontend
 This will Run graph.json on server.
 That is, you need upload graph.json first.
 
-![](./images/run.webp)
+![](./images/run_window.webp)
 
 ## Run
-Execute graph.json into LangGraph with Ollama.
-After Run, recommand download files.
+Execute graph.json into LangGraph with LLM.
+After Run, recommand get files from server.
 
-## Stop
-To terminal remote running server. (Currently, many bugs. If stuck, plese restart docker compose)
-
-## Cancel
+## Leave
 Leave run menu
 
 ## Text Space

Frontend/UI.md

@@ -0,0 +1,29 @@
+---
+layout: frontend
+
+---
+
+
+# UI
+
+There are several buttoms  
+![](./images/panel_major.webp)
+
+## New Graph
+
+New will clean the canvas.
+
+## Save Graph
+
+Save the graph into json to your computer.
+
+## Load Graph
+Read the graph json from your computer.
+
+## Run Graph
+Run graph.json on server, for more info, see [Run](./Run)
+
+
+## Config
+* set what llm you want to use
+* api key

Frontend/UI1.md

@@ -0,0 +1,21 @@
+---
+layout: frontend
+
+---
+
+
+# UI
+
+There are several buttoms  
+![](./images/panel_server.webp)
+
+## Upload Files to Server
+
+Send file to workspace as environment, for such file processing.
+
+## Get Files from Server
+
+After process with LLM, download all files in workspace.
+
+## Clean Server Cache
+Remove all files at workspace