xkerser's picture
Update app.py
db422e9 verified
#!/usr/bin/env python
# coding=utf-8
# Copyright 2024 The Footscray Coding Collective. All rights reserved.
"""
Financial Research Agent: Advanced Market Analysis and Data Access
This script implements a comprehensive financial research agent capable of performing market analysis,
retrieving financial data, and providing interactive research capabilities through either a GUI or
command-line interface.
The agent leverages the Smolagents framework to create an autonomous system that can:
1. Access and analyze real-time market data through Alpha Vantage API integration
2. Process financial documents and extract relevant information
3. Perform web searches and analyze webpage content
4. Create visualizations of financial data
5. Generate comprehensive financial analysis reports
6. Handle user uploads of various document types
Key Components:
-------------
- ModelManager: Handles loading and configuration of various LLM models
- ToolRegistry: Manages initialization and organization of tools available to the agent
- GradioUI: Provides a user-friendly interface with responsive design for desktop/mobile
- A robust set of financial tools for retrieving stock data, financial statements, and market sentiment
- Web browsing capabilities with text extraction and analysis
- Document processing for PDFs, spreadsheets, and other common file formats
- Visualization tools for creating charts and graphs from financial data
Usage:
-----
Run in UI mode (default):
python app.py
Run in headless mode with a specific query:
python app.py --mode headless --query "Analyze Tesla's financial performance for 2023"
Configuration:
------------
The script uses environment variables for API keys and other configuration settings.
Required environment variables:
- ALPHA_VANTAGE_API_KEY: For accessing financial data APIs
- HF_TOKEN: For accessing Hugging Face models (optional)
The agent also maintains detailed logs in the logs/ directory for debugging and auditing.
Dependencies:
-----------
- smolagents: Core framework for agent capabilities
- gradio: For the web interface
- Alpha Vantage API integration: For financial data
- Various data processing libraries: For handling and analyzing financial information
Technical Notes:
--------------
- The agent runs with a configurable number of maximum steps (default: 20)
- Planning occurs at regular intervals (default: every 4 steps)
- The agent has access to a curated list of authorized Python imports for security
- All file uploads are validated for type and size before processing
Created by the Footscray Coding Collective
Copyright 2024, All rights reserved
"""
import contextlib
import datetime
import logging
import mimetypes
import os
import re
import shutil
from typing import Any, Dict, Generator, List, Optional, Tuple
# Typer for CLI functionality
import typer
# Telemetry imports (optional)
# with contextlib.suppress(ImportError):
# from openinference.instrumentation.smolagents import SmolagentsInstrumentor
# from phoenix.otel import register
# Initialize telemetry for observability and tracing
# register()
# SmolagentsInstrumentor().instrument()
# third-party
import gradio as gr
import pytz
from dotenv import load_dotenv
from huggingface_hub import login
from rich.console import Console
from rich.logging import RichHandler
from smolagents import FinalAnswerTool # smolagents
from smolagents import (
CodeAgent,
GoogleSearchTool,
HfApiModel,
LiteLLMModel,
OpenAIServerModel,
Tool,
TransformersModel,
)
from smolagents.agent_types import AgentText, handle_agent_output_types
from smolagents.gradio_ui import pull_messages_from_step
# local
from scripts.finance_tools import (
DataVisualizationTool,
FinancialCalculatorTool,
TrendAnalysisTool,
get_balance_sheet_data,
get_cash_flow_data,
get_company_overview_data,
get_earnings_data,
get_income_statement_data,
get_market_news_sentiment,
get_stock_quote_data,
get_time_series_daily,
search_symbols,
)
from scripts.flux_lora_tool import FluxLoRATool
from scripts.text_cleaner_tool import TextCleanerTool
from scripts.text_inspector_tool import TextInspectorTool
from scripts.text_web_browser import (
ArchiveSearchTool,
DownloadTool,
FinderTool,
FindNextTool,
PageDownTool,
PageUpTool,
SimpleTextBrowser,
VisitTool,
)
from scripts.time_tools import get_temporal_context
from scripts.visual_qa import visualizer
# Initialize console and app
console = Console()
app = typer.Typer(
help="Financial Research Agent - Access market data and analysis through a CLI or UI",
add_completion=False,
)
# ------------------------ Configuration and Setup ------------------------
# Constants and configurations
AUTHORIZED_IMPORTS = [
"requests", # Web requests (fetching data from the internet)
"pytz", # Timezone handling
"zipfile", # Working with ZIP archives
"pandas", # Data manipulation and analysis (DataFrames)
"numpy", # Numerical computing (arrays, linear algebra)
"sympy", # Symbolic mathematics (algebra, calculus)
"json", # JSON data serialization/deserialization
"bs4", # Beautiful Soup for HTML/XML parsing
"pubchempy", # Accessing PubChem chemical database
"yaml",
"xml", # XML processing
"yahoo_finance", # Fetching stock datauv
"Bio", # Bioinformatics tools (e.g., sequence analysis)
"sklearn", # Scikit-learn for machine learning
"scipy", # Scientific computing (stats, optimization)
"pydub", # Audio manipulation
"PIL", # Pillow for image processing
"chess", # Chess-related functionality
"PyPDF2", # PDF manipulation
"pptx", # PowerPoint file manipulation
"torch", # PyTorch for neural networks
"datetime", # Date and time handling
"fractions", # Rational number arithmetic
"csv", # CSV file reading/writing
"cleantext", # Text cleaning and normalization
"os", # Operating system interaction (file system, etc.) VERY IMPORTANT
"re", # Regular expressions for text processing
"collections", # Useful data structures (e.g., defaultdict, Counter)
"math", # Basic mathematical functions
"random", # Random number generation
"io", # Input/output streams
"urllib.parse", # URL parsing and manipulation (safe URL handling)
"typing", # Support for type hints (improve code clarity)
"concurrent.futures", # For parallel execution
"time", # Measuring time
"tempfile", # Creating temporary files and directories
# Data Visualization (if needed) - Consider security implications carefully
"matplotlib.plt", # Plotting library
"seaborn", # Statistical data visualization (more advanced)
# Web Scraping (more specific/controlled) - Consider ethical implications
"lxml", # Faster XML/HTML processing (alternative to bs4)
"selenium", # Automated browser control (for dynamic websites)
# Database interaction (if needed) - Handle credentials securely!
"sqlite3", # SQLite database access
# Task scheduling
"schedule", # Allow the agent to schedule tasks
"uuid",
"base64",
"smolagents", # smolagents package to be able to create smolagents tools
]
USER_AGENT = (
"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 "
"(KHTML, like Gecko) Chrome/119.0.0.0 Safari/537.36 Edg/119.0.0.0"
)
BROWSER_CONFIG = {
"viewport_size": 1024 * 5,
"downloads_folder": "data/downloads_folder",
"request_kwargs": {
"headers": {"User-Agent": USER_AGENT},
"timeout": 300,
},
"serpapi_key": os.getenv("SERPAPI_API_KEY"),
}
CUSTOM_ROLE_CONVERSIONS = {"tool-call": "assistant", "tool-response": "user"}
ALLOWED_FILE_TYPES = [
"application/pdf",
"application/vnd.openxmlformats-officedocument.wordprocessingml.document",
"text/plain",
"text/markdown",
"application/json",
"image/png",
"image/webp",
"image/jpeg",
"image/gif",
"video/mp4",
"audio/mpeg",
"audio/wav",
"audio/ogg",
]
# Set up logging configuration
def setup_logging() -> Tuple[str, logging.Logger]:
"""
Configure logging with structured output and file storage.
The function creates logs directory and timestamped log filename, sets up
logging with Rich integration and creates and returns logger.
Returns:
Tuple containing the log file path and configured logger
"""
# Create logs directory
current_dir = os.path.dirname(os.path.abspath(__file__))
logs_dir = os.path.join(current_dir, "logs")
os.makedirs(logs_dir, exist_ok=True)
# Generate timestamped log filename
melbourne_timezone = pytz.timezone("Australia/Melbourne")
log_filename = f'smolagents_{datetime.datetime.now(melbourne_timezone).strftime("%Y%m%d_%H%M%S")}.log'
log_file = os.path.join(logs_dir, log_filename)
# Set up logging with Rich integration
logging.basicConfig(
level=logging.INFO,
format="%(asctime)s [%(levelname)s] - %(message)s",
datefmt="%Y-%m-%d %H:%M:%S",
handlers=[
RichHandler(rich_tracebacks=True, show_time=True),
logging.FileHandler(log_file),
],
)
# Create and return logger
logger = logging.getLogger(__name__)
return log_file, logger
LOG_FILE, logger = setup_logging()
def setup_environment() -> None:
"""Initialize environment variables and authentication.
This function ensures that required environment variables are set and
attempts to authenticate with Hugging Face and Alpha Vantage services.
"""
load_dotenv(override=True)
# Check Hugging Face token
if os.getenv("HF_TOKEN"): # Check if token is actually set
login(os.getenv("HF_TOKEN"))
console.print("HF_TOKEN loaded successfully")
else:
console.print(
"[yellow]HF_TOKEN not found in environment variables. "
"Some features may not work properly.[/yellow]"
)
# Check Alpha Vantage API key
try:
# Ensure Alpha Vantage API key is available
api_key = os.getenv("ALPHA_VANTAGE_API_KEY")
if not api_key:
console.print(
"[yellow]⚠️ Warning: ALPHA_VANTAGE_API_KEY not found. "
"Finance tools may not work properly.[/yellow]"
)
else:
console.print("[green]✓ ALPHA_VANTAGE_API_KEY loaded successfully[/green]")
except Exception as e:
console.print(f"[red]Error checking ALPHA_VANTAGE_API_KEY: {e}[/red]")
# ------------------------ Model and Tool Management ------------------------
class ModelManager:
"""Manages model loading and initialization.
This class provides a static method to load the specified model with the
appropriate configuration. It supports the following inference types:
- hf_api: Use the Hugging Face API to load the model.
- hf_api_provider: Use the Hugging Face API to load the model with the
'together' provider.
- litellm: Load the LiteLLM model with the specified model ID.
- openai: Load the OpenAI model with the specified model ID and API key.
- transformers: Load the Hugging Face transformers model with the
specified model ID and configuration.
"""
@staticmethod
def load_model(chosen_inference: str, model_id: str, key_manager=None):
"""Load the specified model with appropriate configuration.
Args:
chosen_inference (str): The inference type to use.
model_id (str): The model ID to load.
key_manager (Optional[KeyManager]): The key manager to use for
loading the model. Required for OpenAI models.
Raises:
ValueError: If the chosen inference type is invalid.
Exception: If an error occurs while loading the model.
"""
try:
if chosen_inference == "hf_api":
return HfApiModel(model_id=model_id)
if chosen_inference == "hf_api_provider":
return HfApiModel(provider="together")
if chosen_inference == "litellm":
return LiteLLMModel(model_id=model_id)
if chosen_inference == "openai":
if not key_manager:
raise ValueError("Key manager required for OpenAI model")
return OpenAIServerModel(
model_id=model_id, api_key=key_manager.get_key("openai_api_key")
)
if chosen_inference == "transformers":
return TransformersModel(
model_id="HuggingFaceTB/SmolLM2-1.7B-Instruct",
device_map="auto",
max_new_tokens=1000,
)
else:
raise ValueError(f"Invalid inference type: {chosen_inference}")
except Exception as e:
console.print(f"[red]✗ Couldn't load model: {e}[/red]")
raise
# ------------------------ Tool Registration ------------------------
class ToolRegistry:
"""Manages tool initialization and organization using Zhou Protocol priorities."""
@staticmethod
def load_information_tools(model, text_limit=30000):
"""
Initialize and return information analysis tools.
This method creates tools for analyzing text from documents, and other sources.
The information tools should be prioritized first in the agent's toolset.
Args:
model: Language model to use for analysis
text_limit: Maximum character length for text summaries
Returns:
List of information analysis tools
"""
return [
TextInspectorTool(model, text_limit),
]
@staticmethod
def load_utility_tools():
"""
Initialize and return utility tools for text cleaning and normalization.
Returns:
List of utility tools
"""
return [
TextCleanerTool(),
]
@staticmethod
def load_time_tools():
"""
Initialize and return time-related tools.
Returns:
List of time-related tools
"""
return [get_temporal_context]
@staticmethod
def load_finance_tools():
"""
Initialize and return financial analysis tools.
Returns:
List of financial tools in priority order
"""
return [
# Analysis tools first (higher priority)
DataVisualizationTool(),
FinancialCalculatorTool(),
TrendAnalysisTool(),
# Data retrieval tools next
search_symbols,
get_stock_quote_data,
get_company_overview_data,
get_earnings_data,
get_income_statement_data,
get_balance_sheet_data,
get_cash_flow_data,
get_time_series_daily,
get_market_news_sentiment,
]
@staticmethod
def load_web_tools(browser, text_limit=20000):
"""
Initialize and return web interaction tools.
Args:
browser: Browser instance for web navigation
text_limit: Maximum character length for text processing
Returns:
List of web tools in priority order
"""
return [
# Search tools first
GoogleSearchTool(provider="serper"),
# Navigation tools next
VisitTool(browser),
DownloadTool(browser),
# Page interaction tools last
PageUpTool(browser),
PageDownTool(browser),
FinderTool(browser),
FindNextTool(browser),
ArchiveSearchTool(browser),
]
@staticmethod
def load_image_generation_tools():
"""
Initialize and return image generation tools.
Returns:
Image generation tool or fallback
"""
try:
return Tool.from_space(
space_id="xkerser/FLUX.1-dev",
name="image_generator",
description="Generates high-quality AgentImage using the FLUX.1-dev model based on text prompts.",
)
except Exception as e:
console.print(
f"[yellow]✗ Couldn't initialize image generation tool: {e}[/yellow]"
)
return FluxLoRATool()
@staticmethod
def load_final_answer_tool():
"""
Return the final answer tool for providing conclusive responses.
Returns:
List containing the final answer tool
"""
return [FinalAnswerTool()]
def create_agent(model_id: str = "openrouter/google/gemini-2.0-flash-001"):
"""
Create a fresh agent instance with properly configured tools.
This function creates a CodeAgent with tools organized by the Zhou Protocol
priority system, ensuring the most relevant tools are considered first.
Args:
model_id: The ID of the model to use for the agent
Returns:
A configured CodeAgent instance
Raises:
RuntimeError: If agent creation fails
"""
try:
# Initialize model with fallback system
model = _load_model_with_fallback(model_id)
# Initialize tools
text_limit = 30000
browser = SimpleTextBrowser(**BROWSER_CONFIG)
# Collect all tools with proper Zhou Protocol prioritization
information_tools = ToolRegistry.load_information_tools(model, text_limit)
utility_tools = ToolRegistry.load_utility_tools()
finance_tools = ToolRegistry.load_finance_tools()
web_tools = ToolRegistry.load_web_tools(browser)
time_tools = ToolRegistry.load_time_tools()
image_generator = ToolRegistry.load_image_generation_tools()
final_answer = ToolRegistry.load_final_answer_tool()
# Combine all tools with information tools prioritized first
all_tools = (
information_tools # Critical information extraction (highest priority)
+ utility_tools # General utility functions
+ finance_tools # Financial analysis capabilities
+ web_tools # Web search and navigation
+ time_tools # Time context tools
+ [visualizer] # Image analysis
+ [image_generator] # Image generation
+ final_answer # Task completion (always last)
)
# Validate tools before creating agent
_validate_tools(all_tools)
return CodeAgent(
model=model,
tools=all_tools,
max_steps=20,
verbosity_level=2,
additional_authorized_imports=AUTHORIZED_IMPORTS,
planning_interval=4,
description="""
This agent assists with comprehensive research and financial analysis. It first analyzes
any provided documents or text, then leverages specialized financial tools and web search
capabilities to provide thorough insights.
QUERY COMPREHENSION FRAMEWORK
Before answering any complex question, apply the Zhou Comprehension Pattern:
1. **Initial Parse**: What is literally being asked?
2. **Intent Detection**: What is the user actually trying to accomplish?
3. **Knowledge Assessment**: What information is needed to address this properly?
4. **Tool Selection**: Which tools provide the most direct path to a solution?
5. **Execution Planning**: What sequence of operations will yield the best result?
CLARIFICATION CHECKLIST
When faced with ambiguous queries, the agent should systematically clarify:
* **Scope**: "How comprehensive should this analysis be?"
* **Format**: "What form would you like the results in?"
* **Technical Level**: "Should I explain technical details or focus on practical applications?"
* **Time Horizon**: "Are you interested in historical data, current status, or future projections?"
* **Priority**: "Which aspect of this question is most important to you?"
""".strip(),
)
except Exception as e:
console.print(f"[red]✗ Agent creation failed: {e}[/red]")
raise RuntimeError(f"Agent creation failed: {e}")
def _load_model_with_fallback(model_id: str) -> Any:
"""
Attempt to load the specified model with fallbacks if it fails.
Args:
model_id: Primary model ID to try loading
Returns:
Loaded model instance
Raises:
RuntimeError: If all model loading attempts fail
"""
# Fallback model chain from most capable to most reliable
fallback_models = [
model_id, # Try the requested model first
"openrouter/anthropic/claude-3.7-sonnet",
"openai/gpt-4o-mini",
"anthropic/claude-3.7-sonnet",
"HuggingFaceTB/SmolLM2-1.7B-Instruct", # Last resort local option
]
last_error = None
for model in fallback_models:
try:
return LiteLLMModel(
custom_role_conversions=CUSTOM_ROLE_CONVERSIONS,
model_id=model,
)
except Exception as e:
last_error = e
console.print(f"[yellow]Failed to load model {model}: {e}[/yellow]")
# If we get here, all models failed
raise RuntimeError(f"All model loading attempts failed. Last error: {last_error}")
def _validate_tools(tools):
"""
Validate that all tools are proper Tool instances.
Args:
tools: List of tools to validate
Raises:
ValueError: If any tool is not a Tool instance
"""
for tool in tools:
if not isinstance(tool, Tool):
raise ValueError(
f"Invalid tool type: {type(tool)}. "
f"All tools must be instances of Tool class."
)
# ------------------------ Gradio UI Components ------------------------
def stream_to_gradio(
agent,
task: str,
reset_agent_memory: bool = False,
additional_args: Optional[dict] = None,
):
"""Streams agent responses with improved status indicators."""
try:
# Initial processing indicator
yield gr.ChatMessage(role="assistant", content="⏳ Processing your request...")
# Track what we've yielded to replace the processing indicator
first_message_yielded = False
for step_log in agent.run(
task, stream=True, reset=reset_agent_memory, additional_args=additional_args
):
# The key fix: pull_messages_from_step is a generator function that yields messages
# We need to iterate through each yielded message
for message in pull_messages_from_step(step_log):
if not first_message_yielded:
# Replace the initial "Processing" message
first_message_yielded = True
message.content = message.content.replace(
"⏳ Processing your request...", ""
)
# Check what type of operation is being performed based on the metadata or content
# Instead of trying to access a 'status' attribute that doesn't exist
content_lower = (
message.content.lower() if hasattr(message, "content") else ""
)
if "document analysis" in content_lower:
message.content = f"📄 **Document Analysis:** {message.content}"
elif "search" in content_lower:
message.content = f"🔍 **Search:** {message.content}"
yield message
# Final answer with enhanced formatting
final_answer = handle_agent_output_types(step_log)
if isinstance(final_answer, AgentText):
yield gr.ChatMessage(
role="assistant",
content=f"✅ **Final Answer:**\n\n{final_answer.to_string()}",
)
else:
yield gr.ChatMessage(
role="assistant", content=f"✅ **Final Answer:** {str(final_answer)}"
)
except Exception as e:
yield gr.ChatMessage(
role="assistant",
content=f"❌ **Error:** {str(e)}\n\nPlease try again with a different query.",
)
# ------------------------ Gradio UI Components ------------------------
class GradioUI:
"""A one-line interface to launch your agent in Gradio."""
def __init__(self, file_upload_folder: str | None = None):
"""Initialize the Gradio UI with optional file upload functionality."""
self.file_upload_folder = file_upload_folder
if self.file_upload_folder is not None:
if not os.path.exists(file_upload_folder):
os.mkdir(file_upload_folder)
def interact_with_agent(
self,
prompt: str,
messages: List[gr.ChatMessage],
session_state: Dict[str, Any],
) -> Generator[List[gr.ChatMessage], None, None]:
"""Main interaction handler with the agent.
Args:
prompt: The user's input prompt
messages: The list of messages so far (including the user's prompt)
session_state: The current state of the user's session
Yields:
A list of messages after each step (including the user's prompt)
"""
# Get or create session-specific agent
if "agent" not in session_state:
model_id = session_state.get(
"model_id", "openrouter/google/gemini-2.0-flash-001"
)
session_state["agent"] = create_agent(model_id)
# Adding monitoring
try:
# Log the existence of agent memory
has_memory = hasattr(session_state["agent"], "memory")
console.print(f"Agent has memory: {has_memory}")
if has_memory:
console.print(f"Memory type: {type(session_state['agent'].memory)}")
messages.append(gr.ChatMessage(role="user", content=prompt))
yield messages
for msg in stream_to_gradio(
session_state["agent"], task=prompt, reset_agent_memory=False
):
messages.append(msg)
yield messages # Yield messages after each step
yield messages # Yield messages one last time
except Exception as e:
console.print(f"[red]Error in interaction: {str(e)}[/red]")
raise
def upload_file(
self,
file,
file_uploads_log,
):
"""Handle file uploads with proper validation and security."""
if file is None:
return gr.Textbox("No file uploaded", visible=True), file_uploads_log
try:
mime_type, _ = mimetypes.guess_type(file.name)
except Exception as e:
return gr.Textbox(f"Error: {e}", visible=True), file_uploads_log
if mime_type not in ALLOWED_FILE_TYPES:
return gr.Textbox("File type disallowed", visible=True), file_uploads_log
# Sanitize file name
original_name = os.path.basename(file.name)
sanitized_name = re.sub(
r"[^\w\-.]", "_", original_name
) # Replace invalid chars with underscores
# Ensure the extension correlates to the mime type
type_to_ext = {}
for ext, t in mimetypes.types_map.items():
if t not in type_to_ext:
type_to_ext[t] = ext
# Build sanitized filename with proper extension
name_parts = sanitized_name.split(".")[:-1]
extension = type_to_ext.get(mime_type, "")
sanitized_name = "".join(name_parts) + extension
# Limit File Size, and Throw Error
max_file_size_mb = 50 # Define the limit
file_size_mb = os.path.getsize(file.name) / (1024 * 1024) # Size in MB
if file_size_mb > max_file_size_mb:
return (
gr.Textbox(
f"File size exceeds {max_file_size_mb} MB limit.", visible=True
),
file_uploads_log,
)
# Save the uploaded file to the specified folder
file_path = os.path.join(self.file_upload_folder, sanitized_name)
shutil.copy(file.name, file_path)
return gr.Textbox(
f"File uploaded: {file_path}", visible=True
), file_uploads_log + [file_path]
def log_user_message(self, text_input, file_uploads_log):
"""Process user message and handle file references."""
message = text_input
if len(file_uploads_log) > 0:
message += f"\nYou have been provided with these files, which might be helpful or not: {file_uploads_log}" # Added file list
return (
message,
gr.Textbox(
value="",
interactive=False,
placeholder="Processing...", # Changed placeholder.
),
gr.Button(interactive=False),
)
def detect_device(self, request: gr.Request):
"""Detect whether the user is on mobile or desktop device."""
if not request:
return "Unknown device" # Handle case where request is none.
# Method 1: Check sec-ch-ua-mobile header
is_mobile_header = request.headers.get("sec-ch-ua-mobile")
if is_mobile_header:
return "Mobile" if "?1" in is_mobile_header else "Desktop"
# Method 2: Check user-agent string
user_agent = request.headers.get("user-agent", "").lower()
mobile_keywords = ["android", "iphone", "ipad", "mobile", "phone"]
if any(keyword in user_agent for keyword in mobile_keywords):
return "Mobile"
# Method 3: Check platform
platform = request.headers.get("sec-ch-ua-platform", "").lower()
if platform:
if platform in ['"android"', '"ios"']:
return "Mobile"
if platform in ['"windows"', '"macos"', '"linux"']:
return "Desktop"
# Default case if no clear indicators
return "Desktop"
def launch(self, **kwargs):
"""Launch the Gradio UI with responsive layout."""
with gr.Blocks(theme="ocean", fill_height=True) as demo:
# Different layouts for mobile and computer devices
@gr.render()
def layout(request: gr.Request):
device = self.detect_device(request)
console.print(f"device - {device}")
# Render layout with sidebar
if device == "Desktop":
return self._create_desktop_layout()
return self._create_mobile_layout()
demo.queue(max_size=20).launch(
debug=True, **kwargs
) # Add queue with reasonable size
def _create_desktop_layout(self):
"""Create the desktop layout with sidebar."""
with gr.Blocks(fill_height=True) as sidebar_demo:
with gr.Sidebar():
gr.Markdown(
"""#OpenDeepResearch - 3theSmolagents!
Model_id: google/gemini-2.0-flash-001"""
)
with gr.Group():
gr.Markdown("**What's on your mind mate?**", container=True)
text_input = gr.Textbox(
lines=3,
label="Your request",
container=False,
placeholder="Enter your prompt here and press Shift+Enter or press the button",
)
launch_research_btn = gr.Button("Run", variant="primary")
# If an upload folder is provided, enable the upload feature
if self.file_upload_folder is not None:
upload_file = gr.File(label="Upload a file")
upload_status = gr.Textbox(
label="Upload Status", interactive=False, visible=False
)
file_uploads_log = gr.State([])
upload_file.change(
self.upload_file,
[upload_file, file_uploads_log],
[upload_status, file_uploads_log],
)
gr.HTML("<br><br><h4><center>Powered by:</center></h4>")
with gr.Row():
gr.HTML(
"""
<div style="display: flex; align-items: center; gap: 8px; font-family: system-ui, -apple-system, sans-serif;">
<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/smolagents/mascot_smol.png"
style="width: 32px; height: 32px; object-fit: contain;" alt="logo">
<a target="_blank" href="https://github.com/huggingface/smolagents">
<b>huggingface/smolagents</b>
</a>
</div>
"""
)
# Add session state to store session-specific data
session_state = gr.State({}) # Initialize empty state for each session
stored_messages = gr.State([])
if "file_uploads_log" not in locals():
file_uploads_log = gr.State([])
chatbot = gr.Chatbot(
label="Research-Assistant",
type="messages",
avatar_images=(
None,
"https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/smolagents/mascot_smol.png",
),
resizeable=False,
scale=1,
elem_id="my-chatbot",
)
self._connect_event_handlers(
text_input,
launch_research_btn,
file_uploads_log,
stored_messages,
chatbot,
session_state,
)
return sidebar_demo
def _create_mobile_layout(self):
"""Create the mobile layout (simpler without sidebar)."""
with gr.Blocks(fill_height=True) as simple_demo:
gr.Markdown("""#OpenDeepResearch - free the AI agents!""")
# Add session state to store session-specific data
session_state = gr.State({})
stored_messages = gr.State([])
file_uploads_log = gr.State([])
chatbot = gr.Chatbot(
label="Research-Assistant",
type="messages",
avatar_images=(
None,
"https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/smolagents/mascot_smol.png",
),
resizeable=True,
scale=1,
)
# If an upload folder is provided, enable the upload feature
if self.file_upload_folder is not None:
upload_file = gr.File(label="Upload a file")
upload_status = gr.Textbox(
label="Upload Status", interactive=False, visible=False
)
upload_file.change(
self.upload_file,
[upload_file, file_uploads_log],
[upload_status, file_uploads_log],
)
text_input = gr.Textbox(
lines=1,
label="What's on your mind mate?",
placeholder="Chuck in a question and we'll take care of the rest",
)
launch_research_btn = gr.Button("Run", variant="primary")
self._connect_event_handlers(
text_input,
launch_research_btn,
file_uploads_log,
stored_messages,
chatbot,
session_state,
)
return simple_demo
def _connect_event_handlers(
self,
text_input,
launch_research_btn,
file_uploads_log,
stored_messages,
chatbot,
session_state,
):
"""Connect the event handlers for input elements."""
# Connect text input submit event
text_input.submit(
self.log_user_message,
[text_input, file_uploads_log],
[stored_messages, text_input, launch_research_btn],
).then(
self.interact_with_agent,
[stored_messages, chatbot, session_state],
[chatbot],
).then(
lambda: (
gr.Textbox(
interactive=True,
placeholder="Enter your prompt here and press the button",
),
gr.Button(interactive=True),
),
None,
[text_input, launch_research_btn],
)
# Connect button click event
launch_research_btn.click(
self.log_user_message,
[text_input, file_uploads_log],
[stored_messages, text_input, launch_research_btn],
).then(
self.interact_with_agent,
[stored_messages, chatbot, session_state],
[chatbot],
).then(
lambda: (
gr.Textbox(
interactive=True,
placeholder="Enter your prompt here and press the button",
),
gr.Button(interactive=True),
),
None,
[text_input, launch_research_btn],
)
# ------------------------ CLI Command ------------------------
@app.command()
def run(
mode: str = typer.Option(
"ui",
"--mode",
"-m",
help="Operating mode: 'ui' for Gradio interface or 'headless' for CLI mode",
),
model_id: str = typer.Option(
"openrouter/google/gemini-2.0-flash-001",
"--model",
help="Model ID to use for the agent",
),
query: Optional[str] = typer.Option(
None, "--query", "-q", help="Query to execute (required in headless mode)"
),
):
"""
Run the financial research agent in either UI or headless mode.
In UI mode, launches a Gradio interface for interactive use.
In headless mode, processes a single query and outputs the result to the console.
"""
# Setup environment variables
setup_environment()
# Validate inputs for headless mode
if mode == "headless" and not query:
console.print("[red]Error: query parameter is required in headless mode[/red]")
raise typer.Exit(code=1)
# Create agent with specified model ID
console.print(f"[bold]Initializing agent with model:[/bold] {model_id}")
# Execute in appropriate mode
if mode == "ui":
console.print(
"[bold green]Starting UI mode with Gradio interface...[/bold green]"
)
# Ensure downloads folder exists
os.makedirs(f"./{BROWSER_CONFIG['downloads_folder']}", exist_ok=True)
# Launch UI
GradioUI(file_upload_folder="data/uploaded_files").launch()
elif mode == "headless":
console.print(f"[bold]Processing query in headless mode:[/bold] {query}")
# Create agent for headless mode
agent = create_agent(model_id)
# Show a simple spinner during processing
with console.status("[bold green]Processing query...[/bold green]"):
result = agent.run(query)
# Display the results
console.print("\n[bold green]Results:[/bold green]")
console.print(result)
else:
console.print(
f"[red]Error: Invalid mode '{mode}'. Use 'ui' or 'headless'[/red]"
)
raise typer.Exit(code=1)
# ------------------------ Main Entry Point ------------------------
if __name__ == "__main__":
# Use the typer app as the entry point
app()