Add integration README and fix linting

Co-authored-by: hiyouga <16256802+hiyouga@users.noreply.github.com>

Author: Copilot

src_py/integration/README.md

@@ -0,0 +1,104 @@
+# Integration Module
+
+This module contains integration tools for AgentHub, including conversation tracing and a web UI for interacting with LLMs.
+
+## Components
+
+### Tracer
+
+The `Tracer` class provides functionality to save conversation history to local files and browse them via a web interface.
+
+**Features:**
+- Save conversation history as JSON and text files
+- Web interface for browsing saved conversations
+- Automatic serialization of complex data types
+
+**Usage:**
+
+```python
+from integration import Tracer
+
+# Initialize tracer
+tracer = Tracer(cache_dir="cache")
+
+# Save conversation history
+history = [
+    {"role": "user", "content_items": [{"type": "text", "text": "Hello"}]},
+    {"role": "assistant", "content_items": [{"type": "text", "text": "Hi there!"}]}
+]
+config = {"temperature": 0.7}
+tracer.save_history(history, "conversation_001", config)
+
+# Start web server to browse saved conversations
+tracer.start_web_server(host="127.0.0.1", port=5000)
+```
+
+**With AutoLLMClient:**
+
+```python
+from agenthub import AutoLLMClient
+
+client = AutoLLMClient(model="gemini-3-flash-preview")
+config = {"trace_id": "agent1/conversation_001", "temperature": 0.7}
+
+message = {"role": "user", "content_items": [{"type": "text", "text": "Hello"}]}
+async for _ in client.streaming_response_stateful(message=message, config=config):
+    pass
+# Conversation is automatically saved to cache/agent1/conversation_001.json and .txt
+```
+
+### Web UI
+
+The Web UI provides an interactive chat interface for communicating with LLMs through a web browser.
+
+**Features:**
+- Interactive chat interface
+- Editable configuration (model, temperature, max_tokens)
+- Streaming message display
+- Message cards showing token usage and finish reason
+- Support for multiple AI models (Gemini, Claude, GLM)
+
+**Usage:**
+
+```python
+from integration import start_chat_server
+
+# Start the chat server
+start_chat_server(host="127.0.0.1", port=5001, debug=False)
+# Open http://127.0.0.1:5001 in your browser
+```
+
+**Or run the example:**
+
+```bash
+python examples/web_ui_example.py
+```
+
+**Features:**
+- **Config Panel**: Click the "⚙️ Config" button to edit:
+  - Model selection (Gemini 3 Flash, Claude Sonnet 4.5, GLM 4.7)
+  - Temperature (0-2)
+  - Max Tokens
+- **Message Input**: Type messages in the input box (Shift+Enter for new line, Enter to send)
+- **Message Cards**: Each message is displayed as a card with:
+  - Role indicator (User/Assistant)
+  - Message content
+  - Token usage (in bottom-right corner)
+  - Finish reason (in bottom-right corner)
+- **Streaming**: Responses stream in real-time as the model generates them
+- **Clear**: Clear the conversation history
+
+## Examples
+
+See the `examples/` directory for usage examples:
+- `trace_example.py` - Demonstrates conversation tracing
+- `web_ui_example.py` - Starts the web UI server
+
+## API Endpoints
+
+The Web UI exposes the following endpoints:
+
+- `GET /` - Main chat interface
+- `POST /api/chat` - Chat API endpoint (accepts streaming requests)
+  - Request body: `{"message": {...}, "config": {...}, "history": [...]}`
+  - Response: Server-Sent Events (SSE) stream with chat events

src_py/tests/test_web_ui.py

@@ -12,7 +12,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import pytest
 from flask import Flask
 
 from integration.web_ui import create_chat_app