Spaces:
Sleeping
Sleeping
File size: 24,018 Bytes
bdedf43 522f7a0 1099afe 522f7a0 bdedf43 522f7a0 1099afe 522f7a0 1099afe 522f7a0 1099afe 522f7a0 1099afe 522f7a0 1099afe 522f7a0 1099afe 522f7a0 1099afe 522f7a0 1099afe 522f7a0 1099afe 522f7a0 1099afe 522f7a0 bdedf43 522f7a0 1099afe 522f7a0 bdedf43 522f7a0 bdedf43 1099afe 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 1099afe 522f7a0 1099afe 522f7a0 1099afe 522f7a0 1099afe 522f7a0 bdedf43 522f7a0 1099afe bdedf43 522f7a0 1099afe 522f7a0 bdedf43 522f7a0 1099afe 522f7a0 1099afe 522f7a0 1099afe 522f7a0 1099afe 522f7a0 1099afe 522f7a0 bdedf43 522f7a0 1099afe 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 1099afe 522f7a0 1099afe 522f7a0 1099afe 522f7a0 1099afe bdedf43 522f7a0 1099afe bdedf43 1099afe bdedf43 1099afe 522f7a0 bdedf43 1099afe 522f7a0 bdedf43 522f7a0 1099afe 522f7a0 bdedf43 522f7a0 1099afe 522f7a0 1099afe bdedf43 1099afe 522f7a0 1099afe bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 1099afe bdedf43 1099afe bdedf43 522f7a0 bdedf43 522f7a0 1099afe bdedf43 1099afe bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 bdedf43 522f7a0 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 |
# DEPENDENCIES
import sys
from typing import Any
from typing import Dict
from typing import List
from pathlib import Path
from typing import Optional
# Add parent directory to path for imports
sys.path.append(str(Path(__file__).parent.parent))
from services.risk_analyzer import RiskScore
from services.data_models import SummaryContext
from utils.logger import ContractAnalyzerLogger
from model_manager.llm_manager import LLMManager
from model_manager.llm_manager import LLMProvider
from services.data_models import ContractCategory
from services.data_models import RiskInterpretation
from services.data_models import NegotiationPlaybook
class SummaryGenerator:
"""
LLM-powered executive summary generator for contract analysis : Generates professional, detailed executive summaries using ALL pipeline outputs
"""
def __init__(self, llm_manager: Optional[LLMManager] = None):
"""
Initialize the summary generator
Arguments:
----------
llm_manager { LLMManager } : LLM manager instance (if None, creates one with default settings)
"""
self.llm_manager = llm_manager or LLMManager()
self.logger = ContractAnalyzerLogger.get_logger()
self.logger.info("Summary generator initialized")
# Main entry point with full pipeline integration
def generate_executive_summary(self, contract_text: str, classification: ContractCategory, risk_analysis: RiskScore, risk_interpretation: RiskInterpretation,
negotiation_playbook: NegotiationPlaybook, unfavorable_terms: List, missing_protections: List, clauses: List) -> str:
"""
Generate executive summary using all the pipeline outputs
Arguments:
----------
contract_text { str } : Original contract text (for context)
classification { ContractCategory } : Contract classification results
risk_analysis { RiskScore } : Complete risk analysis
risk_interpretation { RiskInterpretation } : LLM-enhanced risk explanations
negotiation_playbook { NegotiationPlaybook } : Comprehensive negotiation strategy
unfavorable_terms { List } : Detected unfavorable terms
missing_protections { List } : Missing protections
clauses { List } : Extracted clauses
Returns:
--------
{ str } : Generated executive summary string
"""
try:
# Prepare context with all pipeline data
context = self._prepare_summary_context(contract_text = contract_text,
classification = classification,
risk_analysis = risk_analysis,
risk_interpretation = risk_interpretation,
negotiation_playbook = negotiation_playbook,
unfavorable_terms = unfavorable_terms,
missing_protections = missing_protections,
clauses = clauses,
)
# Generate summary using LLM
summary = self._generate_summary(context = context)
self.logger.info(f"Executive summary generated - Risk: {context.risk_score}/100 ({context.risk_level})")
return summary
except Exception as e:
self.logger.error(f"Failed to generate comprehensive summary: {repr(e)}")
# Fallback with available data
return self._generate_fallback_summary(contract_text = contract_text,
classification = classification,
risk_analysis = risk_analysis,
unfavorable_terms = unfavorable_terms,
missing_protections = missing_protections,
)
def _prepare_summary_context(self, contract_text: str, classification: ContractCategory, risk_analysis: RiskScore, risk_interpretation: RiskInterpretation,
negotiation_playbook: NegotiationPlaybook, unfavorable_terms: List[Dict], missing_protections: List[Dict], clauses: List) -> SummaryContext:
"""
Prepare summary context with all pipeline data
"""
# Handle null negotiation_playbook
walk_away_count = 0
if negotiation_playbook and hasattr(negotiation_playbook, 'walk_away_items'):
walk_away_count = len(negotiation_playbook.walk_away_items)
# Extract contract text
contract_preview = contract_text
# Extract key findings from all sources
key_findings = self._extract_findings(risk_analysis = risk_analysis,
risk_interpretation = risk_interpretation,
negotiation_playbook = negotiation_playbook,
unfavorable_terms = unfavorable_terms,
missing_protections = missing_protections,
clauses = clauses,
)
# Prepare metadata
metadata = {"contract_length" : len(contract_text),
"clauses_analyzed" : len(clauses),
"critical_issues" : len([t for t in unfavorable_terms if (self._get_severity(t) == "critical")]),
"walk_away_items" : walk_away_count,
}
return SummaryContext(contract_type = classification.category,
risk_score = risk_analysis.overall_score,
risk_level = risk_analysis.risk_level,
category_scores = risk_analysis.category_scores,
unfavorable_terms = unfavorable_terms,
missing_protections = missing_protections,
clauses = clauses,
key_findings = key_findings,
risk_interpretation = risk_interpretation,
negotiation_playbook = negotiation_playbook,
contract_text_preview = contract_preview,
contract_metadata = metadata,
)
def _extract_findings(self, risk_analysis: RiskScore, risk_interpretation: RiskInterpretation, negotiation_playbook: NegotiationPlaybook,
unfavorable_terms: List[Dict], missing_protections: List[Dict], clauses: List) -> List[str]:
"""
Extract findings from all analysis components
"""
findings = list()
# Overall risk context
if (risk_analysis.overall_score >= 80):
findings.append("CRITICAL RISK LEVEL: Contract presents unacceptable risk requiring immediate attention")
elif (risk_analysis.overall_score >= 60):
findings.append("HIGH RISK LEVEL: Significant concerns requiring substantial negotiation")
# Critical unfavorable terms
critical_terms = [t for t in unfavorable_terms if (self._get_severity(t) == "critical")]
if critical_terms:
findings.append(f"{len(critical_terms)} CRITICAL unfavorable terms identified")
for term in critical_terms[:2]:
term_name = self._get_term_name(term = term)
findings.append(f"Critical: {term_name}")
# Critical missing protections
critical_protections = [p for p in missing_protections if (self._get_importance(p) == "critical")]
if critical_protections:
findings.append(f"{len(critical_protections)} CRITICAL protections missing")
for prot in critical_protections[:2]:
prot_name = self._get_protection_name(protection = prot)
findings.append(f"Missing: {prot_name}")
# High-risk categories
high_risk_categories = [cat for cat, score in risk_analysis.category_scores.items() if (score >= 70)]
if high_risk_categories:
findings.append(f"High-risk categories: {', '.join(high_risk_categories)}")
# Walk-away items from negotiation playbook
if negotiation_playbook and negotiation_playbook.walk_away_items:
findings.append(f"{len(negotiation_playbook.walk_away_items)} potential deal-breakers identified")
# Key concerns from risk interpretation
if risk_interpretation and risk_interpretation.key_concerns:
top_concerns = risk_interpretation.key_concerns[:2]
for concern in top_concerns:
findings.append(f"Key concern: {concern}")
return findings
def _generate_summary(self, context: SummaryContext) -> str:
"""
Generate enhanced summary using comprehensive context
"""
prompt = self._build_summary_prompt(context)
system_prompt = self._build_system_prompt()
try:
response = self.llm_manager.complete(prompt = prompt,
system_prompt = system_prompt,
temperature = 0.3,
max_tokens = 500,
json_mode = False,
)
if response.success and response.text.strip():
return self._clean_summary_response(text = response.text)
else:
raise ValueError(f"LLM generation failed: {response.error_message}")
except Exception as e:
self.logger.error(f"Enhanced LLM summary generation failed: {e}")
# Fallback to basic summary
return self._generate_fallback_summary_from_context(context = context)
def _build_system_prompt(self) -> str:
"""
Build system prompt for executive summary generation
"""
system_prompt = """
You are a senior contract risk analyst. Generate CONCISE executive summaries.
CRITICAL REQUIREMENTS:
1. Maximum 120 words (strict limit)
2. Must mention SPECIFIC clause numbers (e.g., Clause 8.2, Clause 9.5)
3. Direct, urgent tone - no hedging or academic language
4. Focus ONLY on top 3 critical risks
STRUCTURE (3-4 sentences total):
Sentence 1: Overall risk assessment with contract type
Sentence 2-3: Top 2-3 critical risks with SPECIFIC clause references
Sentence 4: Brief actionable conclusion
TONE EXAMPLES:
β
GOOD: "This employment agreement is heavily skewed in favor of the Employer. Clause 8.2 fails to define post-probation salary. Clause 11.2 allows illegal wage forfeiture."
β BAD: "The comprehensive analysis indicates that there are several concerns that require attention. It is essential to carefully review..."
FORBIDDEN PHRASES:
- "comprehensive analysis"
- "it is essential to"
- "requires attention"
- "should be reviewed"
- "it is recommended"
OUTPUT: Pure paragraph text only. No formatting, no bullets, no headers.
"""
return system_prompt
def _build_summary_prompt(self, context: SummaryContext) -> str:
"""
Build prompt for executive summary generation
"""
# Extract top critical issues only
critical_terms = [t for t in context.unfavorable_terms if self._get_severity(t) == "critical"][:10]
critical_protections = [p for p in context.missing_protections if self._get_importance(p) == "critical"][:10]
# Build concise context
critical_issues_text = ""
if critical_terms:
critical_issues_text += "CRITICAL UNFAVORABLE TERMS:\n"
for term in critical_terms:
clause_reference = self._get_clause_reference(term = term)
term_name = self._get_term_name(term = term)
critical_issues_text += f"- {clause_reference}: {term_name}\n"
if critical_protections:
critical_issues_text += "\nCRITICAL MISSING PROTECTIONS:\n"
for protection in critical_protections:
protection_name = self._get_protection_name(protection = protection)
critical_issues_text += f"- {protection_name}\n"
# Determine risk tone
if (context.risk_score >= 80):
risk_tone = "heavily skewed/very high risk/presents unacceptable risk"
elif (context.risk_score >= 60):
risk_tone = "significantly unfavorable/high risk/substantial concerns"
elif (context.risk_score >= 40):
risk_tone = "moderately concerning/notable risk/requires negotiation"
else:
risk_tone = "generally reasonable/manageable risk/standard concerns"
summary_prompt = f"""
CONTRACT ANALYSIS DATA:
- Type: {context.contract_type.replace('_', ' ').title()}
- Risk Score: {context.risk_score}/100
- Risk Level: {context.risk_level}
- Appropriate Tone: {risk_tone}
{critical_issues_text}
TASK:
Write a 100-120 word executive summary following this EXACT structure:
1. First sentence: "This [contract type] [risk assessment with tone matching score]"
2. Second sentence: State top critical risk with SPECIFIC clause number
3. Third sentence: State second critical risk with SPECIFIC clause number
4. Fourth sentence: Brief conclusion about action needed
EXAMPLE (for 85/100 risk employment contract):
"This employment agreement is heavily skewed in favor of the Employer, presenting a very high risk to the Employee. Key concerns include Clause 9.5's extremely broad 24-month non-compete against the entire industry, and Clause 11.2's punitive penalty allowing forfeiture of earned wages. The termination clauses in Clause 17 are highly asymmetrical, giving the employer unilateral power. Significant negotiation is required before signing."
YOUR TURN - Generate summary for THIS contract:
"""
return summary_prompt
def _clean_summary_response(self, text: str) -> str:
"""
Clean and format the LLM response
"""
# Remove any markdown formatting
text = text.replace('**', '').replace('*', '').replace('#', '')
# Remove common LLM artifacts and empty lines
lines = text.split('\n')
cleaned_lines = list()
for line in lines:
line = line.strip()
if line and not line.lower().startswith(('executive summary', 'summary:', 'here is', 'based on', 'certainly')):
cleaned_lines.append(line)
# Join into coherent paragraph
summary = ' '.join(cleaned_lines)
# Ensure proper sentence structure
if summary:
if not summary[0].isupper():
summary = summary[0].upper() + summary[1:]
if not summary.endswith(('.', '!', '?')):
summary += '.'
return summary
def _generate_fallback_summary(self, contract_text: str, classification: ContractCategory, risk_analysis: RiskScore, unfavorable_terms: List[Dict], missing_protections: List[Dict]) -> str:
"""
Generate enhanced fallback summary
"""
contract_type_display = classification.category.replace('_', ' ').title()
# Count critical items
critical_terms = len([t for t in unfavorable_terms if (self._get_severity(t) == "critical")])
critical_protections = len([p for p in missing_protections if (self._get_importance(p) == "critical")])
# Risk assessment
if (risk_analysis.overall_score >= 80):
risk_assessment = f"This {contract_type_display} presents a CRITICAL level of risk"
action = "requires immediate executive attention and significant revision before consideration"
elif (risk_analysis.overall_score >= 60):
risk_assessment = f"This {contract_type_display} presents a HIGH level of risk"
action = "requires careful legal review and substantial negotiation to mitigate key concerns"
elif (risk_analysis.overall_score >= 40):
risk_assessment = f"This {contract_type_display} presents a MODERATE level of risk"
action = "requires professional review and selective negotiation on specific provisions"
else:
risk_assessment = f"This {contract_type_display} presents a LOW level of risk"
action = "appears generally reasonable but should undergo standard legal review"
summary = f"{risk_assessment} with an overall risk score of {risk_analysis.overall_score}/100. "
summary += f"The agreement {action}. "
# Add critical items context
if (critical_terms > 0):
summary += f"Analysis identified {critical_terms} critical unfavorable terms "
if critical_protections > 0:
summary += f"and {critical_protections} critical missing protections. "
else:
summary += f"and {len(missing_protections)} missing standard protections. "
else:
summary += f"Review identified {len(unfavorable_terms)} areas for improvement. "
# Add high-risk categories context
high_risk_categories = [cat for cat, score in risk_analysis.category_scores.items() if (score >= 60)]
if high_risk_categories:
category_names = [cat.replace('_', ' ').title() for cat in high_risk_categories[:2]]
summary += f"Particular attention should be given to {', '.join(category_names)} provisions. "
summary += "Proceed with the detailed negotiation strategy and risk mitigation recommendations provided in the full analysis."
return summary
def _generate_fallback_summary_from_context(self, context: SummaryContext) -> str:
"""
Generate fallback summary from context object
"""
# Access attributes safely, providing defaults if needed by the fallback logic
text_preview = context.contract_text_preview if context.contract_text_preview is not None else ""
missing_prots = context.missing_protections if context.missing_protections is not None else []
unfav_terms = context.unfavorable_terms if context.unfavorable_terms is not None else []
return self._generate_fallback_summary(contract_text = text_preview,
classification = type('MockClassification', (), {'category': context.contract_type})(),
risk_analysis = type('MockRiskAnalysis', (), {'overall_score': context.risk_score, 'risk_level': context.risk_level, 'category_scores': context.category_scores or {}})(),
unfavorable_terms = unfav_terms,
missing_protections = missing_prots,
)
def _get_severity(self, term) -> str:
"""
Safely get severity from term object or dict
"""
try:
if (hasattr(term, 'severity')):
return term.severity
else:
return term.get('severity', 'unknown')
except (AttributeError, KeyError):
return 'unknown'
def _get_importance(self, protection) -> str:
"""
Safely get importance from protection object or dict
"""
try:
if hasattr(protection, 'importance'):
return protection.importance
else:
return protection.get('importance', 'unknown')
except (AttributeError, KeyError):
return 'unknown'
def _get_term_name(self, term) -> str:
"""
Safely get term name
"""
try:
if hasattr(term, 'term'):
return term.term
else:
return term.get('term', 'Unknown Term')
except (AttributeError, KeyError):
return 'Unknown Term'
def _get_protection_name(self, protection) -> str:
"""
Safely get protection name
"""
try:
if hasattr(protection, 'protection'):
return protection.protection
else:
return protection.get('protection', 'Unknown Protection')
except (AttributeError, KeyError):
return 'Unknown Protection'
def _get_explanation(self, item) -> str:
"""
Safely get explanation
"""
try:
if hasattr(item, 'explanation'):
return item.explanation
else:
return item.get('explanation', 'No explanation available')
except (AttributeError, KeyError):
return 'No explanation available'
def _get_clause_reference(self, term) -> str:
"""
Safely get clause reference from term
"""
try:
if hasattr(term, 'clause_reference'):
ref = term.clause_reference
return ref if ref and ref != 'None' else 'Multiple clauses'
else:
ref = term.get('clause_reference', '')
return ref if ref and ref != 'None' else 'Multiple clauses'
except (AttributeError, KeyError):
return 'Unknown clause' |