File size: 12,889 Bytes
499796e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 |
"""
Modal.com Deployment Configuration for Spend Analyzer MCP
"""
import modal
import os
from typing import Dict, Any, Optional
import json
import asyncio
from datetime import datetime
import logging
# Create Modal app
app = modal.App("spend-analyzer-mcp")
# Define the container image with all dependencies
image = (
modal.Image.debian_slim(python_version="3.11")
.pip_install([
"fastapi",
"uvicorn",
"gradio",
"pandas",
"numpy",
"PyPDF2",
"PyMuPDF",
"anthropic",
"python-multipart",
"aiofiles",
"python-dotenv",
"imaplib2",
"email-validator",
"pydantic",
"websockets",
"asyncio-mqtt"
])
.apt_install(["tesseract-ocr", "tesseract-ocr-eng"])
)
# Secrets for API keys and email credentials
secrets = [
modal.Secret.from_name("anthropic-api-key"), # ANTHROPIC_API_KEY
modal.Secret.from_name("email-credentials"), # EMAIL_USER, EMAIL_PASS, IMAP_SERVER
]
# Shared volume for persistent storage
volume = modal.Volume.from_name("spend-analyzer-data", create_if_missing=True)
@app.function(
image=image,
secrets=secrets,
volumes={"/data": volume},
timeout=300,
memory=2048,
cpu=2.0
)
def process_bank_statements(email_config: Dict, days_back: int = 30, passwords: Optional[Dict] = None):
"""
Modal function to process bank statements from email
"""
import sys
sys.path.append("/data")
from email_processor import EmailProcessor, PDFProcessor
from spend_analyzer import SpendAnalyzer
try:
# Initialize processors
email_processor = EmailProcessor(email_config)
pdf_processor = PDFProcessor()
analyzer = SpendAnalyzer()
# Fetch emails
emails = asyncio.run(email_processor.fetch_bank_emails(days_back))
all_transactions = []
processed_statements = []
for email_msg in emails:
try:
# Extract attachments
attachments = asyncio.run(email_processor.extract_attachments(email_msg))
for filename, content, file_type in attachments:
if file_type == 'pdf':
# Try to process PDF
password = None
if passwords and filename in passwords:
password = passwords[filename]
try:
statement_info = asyncio.run(pdf_processor.process_pdf(content, password))
all_transactions.extend(statement_info.transactions)
processed_statements.append({
'filename': filename,
'bank': statement_info.bank_name,
'account': statement_info.account_number,
'period': statement_info.statement_period,
'transaction_count': len(statement_info.transactions)
})
except ValueError as e:
if "password" in str(e).lower():
# PDF requires password
processed_statements.append({
'filename': filename,
'status': 'password_required',
'error': str(e)
})
else:
processed_statements.append({
'filename': filename,
'status': 'error',
'error': str(e)
})
except Exception as e:
logging.error(f"Error processing email: {e}")
continue
# Analyze transactions
if all_transactions:
analyzer.load_transactions(all_transactions)
analysis_data = analyzer.export_analysis_data()
else:
analysis_data = {'message': 'No transactions found'}
return {
'processed_statements': processed_statements,
'total_transactions': len(all_transactions),
'analysis': analysis_data,
'timestamp': datetime.now().isoformat()
}
except Exception as e:
logging.error(f"Error in process_bank_statements: {e}")
return {'error': str(e)}
@app.function(
image=image,
secrets=secrets,
timeout=60
)
def analyze_uploaded_statements(pdf_contents: Dict[str, bytes], passwords: Optional[Dict] = None):
"""
Modal function to analyze directly uploaded PDF statements
"""
from pdf_processor import PDFProcessor
from spend_analyzer import SpendAnalyzer
try:
pdf_processor = PDFProcessor()
analyzer = SpendAnalyzer()
all_transactions = []
processed_files = []
for filename, content in pdf_contents.items():
try:
password = passwords.get(filename) if passwords else None
statement_info = asyncio.run(pdf_processor.process_pdf(content, password))
all_transactions.extend(statement_info.transactions)
processed_files.append({
'filename': filename,
'bank': statement_info.bank_name,
'account': statement_info.account_number,
'transaction_count': len(statement_info.transactions),
'status': 'success'
})
except Exception as e:
processed_files.append({
'filename': filename,
'status': 'error',
'error': str(e)
})
# Analyze transactions
if all_transactions:
analyzer.load_transactions(all_transactions)
analysis_data = analyzer.export_analysis_data()
else:
analysis_data = {'message': 'No transactions found'}
return {
'processed_files': processed_files,
'total_transactions': len(all_transactions),
'analysis': analysis_data
}
except Exception as e:
return {'error': str(e)}
@app.function(
image=image,
secrets=secrets,
volumes={"/data": volume},
timeout=30
)
def get_claude_analysis(analysis_data: Dict, user_question: str = ""):
"""
Modal function to get Claude's analysis of spending data
"""
import anthropic
try:
client = anthropic.Anthropic(api_key=os.environ["ANTHROPIC_API_KEY"])
# Prepare context for Claude
context = f"""
Financial Analysis Data:
{json.dumps(analysis_data, indent=2, default=str)}
User Question: {user_question if user_question else "Please provide a comprehensive analysis of my spending patterns and recommendations."}
"""
response = client.messages.create(
model="claude-3-sonnet-20240229",
max_tokens=1500,
messages=[
{
"role": "user",
"content": f"""
You are a financial advisor analyzing bank statement data.
Based on the provided financial data, give insights about:
1. Spending patterns and trends
2. Budget adherence and alerts
3. Unusual transactions that need attention
4. Specific recommendations for improvement
5. Answer to the user's specific question if provided
Be specific, actionable, and highlight both positive aspects and areas for improvement.
{context}
"""
}
]
)
return {
'claude_analysis': response.content[0].text,
'usage': response.usage.input_tokens + response.usage.output_tokens
}
except Exception as e:
return {'error': f"Claude API error: {str(e)}"}
@app.function(
image=image,
volumes={"/data": volume},
timeout=30
)
def save_user_data(user_id: str, data: Dict):
"""
Save user analysis data to persistent storage
"""
try:
import json
import os
user_dir = f"/data/users/{user_id}"
os.makedirs(user_dir, exist_ok=True)
# Save analysis data
with open(f"{user_dir}/analysis.json", "w") as f:
json.dump(data, f, indent=2, default=str)
# Save timestamp
with open(f"{user_dir}/last_updated.txt", "w") as f:
f.write(datetime.now().isoformat())
return {"status": "saved", "path": user_dir}
except Exception as e:
return {"error": str(e)}
@app.function(
image=image,
volumes={"/data": volume},
timeout=30
)
def load_user_data(user_id: str):
"""
Load user analysis data from persistent storage
"""
try:
import json
user_dir = f"/data/users/{user_id}"
analysis_file = f"{user_dir}/analysis.json"
if os.path.exists(analysis_file):
with open(analysis_file, "r") as f:
data = json.load(f)
# Get last updated time
last_updated = None
if os.path.exists(f"{user_dir}/last_updated.txt"):
with open(f"{user_dir}/last_updated.txt", "r") as f:
last_updated = f.read().strip()
return {
"data": data,
"last_updated": last_updated,
"status": "found"
}
else:
return {"status": "not_found"}
except Exception as e:
return {"error": str(e)}
# Webhook endpoint for MCP integration
@app.function(
image=image,
secrets=secrets,
volumes={"/data": volume}
)
@modal.web_endpoint(method="POST")
def mcp_webhook(request_data: Dict):
"""
Webhook endpoint for MCP protocol messages
"""
try:
from mcp_server import MCPServer
# Initialize MCP server
server = MCPServer()
# Register tools
async def process_statements_tool(args):
email_config = args.get('email_config', {})
days_back = args.get('days_back', 30)
passwords = args.get('passwords', {})
result = process_bank_statements.remote(email_config, days_back, passwords)
return result
async def analyze_pdf_tool(args):
pdf_contents = args.get('pdf_contents', {})
passwords = args.get('passwords', {})
result = analyze_uploaded_statements.remote(pdf_contents, passwords)
return result
async def get_analysis_tool(args):
analysis_data = args.get('analysis_data', {})
user_question = args.get('user_question', '')
result = get_claude_analysis.remote(analysis_data, user_question)
return result
# Register tools with MCP server
server.register_tool("process_email_statements", "Process bank statements from email", process_statements_tool)
server.register_tool("analyze_pdf_statements", "Analyze uploaded PDF statements", analyze_pdf_tool)
server.register_tool("get_claude_analysis", "Get Claude's financial analysis", get_analysis_tool)
# Handle MCP message
response = asyncio.run(server.handle_message(request_data))
return response
except Exception as e:
return {
"jsonrpc": "2.0",
"id": request_data.get("id"),
"error": {
"code": -32603,
"message": str(e)
}
}
# CLI for local testing
@app.local_entrypoint()
def main():
"""
Local entrypoint for testing Modal functions
"""
print("Testing Modal deployment...")
# Test basic functionality
test_data = {
"spending_insights": [],
"recommendations": ["Test recommendation"]
}
result = get_claude_analysis.remote(test_data, "What do you think about my spending?")
print("Claude analysis result:", result)
if __name__ == "__main__":
# For running locally
modal.run(main) |