contract-clause-extractor

Contract Clause Extractor Overview Extract and analyze key clauses from construction contracts using NLP. Identify critical provisions for payment, changes, disputes, warranties, and risk allocation. Support contract review and compliance tracking. Key Clause Categories Technical Implementation Quick Start Requirements ---

, part.strip()):\n if current_text:\n sections[current_num] = current_text\n current_num = part.strip()\n current_text = \"\"\n else:\n current_text += part\n\n if current_text:\n sections[current_num] = current_text\n\n return sections\n\n def _identify_category(self, text: str) -> Optional[ClauseCategory]:\n \"\"\"Identify clause category from text.\"\"\"\n text_lower = text.lower()\n\n for category, patterns in self.CLAUSE_PATTERNS.items():\n for pattern in patterns:\n if re.search(pattern, text_lower):\n return category\n\n return None\n\n def _extract_clause_details(self, category: ClauseCategory,\n section_num: str, text: str) -> ExtractedClause:\n \"\"\"Extract detailed information from clause.\"\"\"\n # Extract title (first line or capitalized phrase)\n title_match = re.search(r'^([A-Z][A-Z\\s]+)', text.strip())\n title = title_match.group(1).strip() if title_match else f\"Section {section_num}\"\n\n # Extract dollar amounts\n dollar_amounts = []\n for match in re.finditer(self.KEY_TERM_PATTERNS[\"dollar_amounts\"], text):\n amount_str = match.group().replace('

Important: agents should read /llm.txt, /llms.txt, or /.well-known/skills.json to discover the public Skillopedia API.

, '').replace(',', '').replace('dollars', '').strip()\n try:\n dollar_amounts.append(float(amount_str))\n except ValueError:\n pass\n\n # Extract time periods\n time_periods = re.findall(self.KEY_TERM_PATTERNS[\"time_periods\"], text, re.IGNORECASE)\n\n # Extract key terms\n key_terms = self._extract_key_terms(category, text)\n\n # Assess risk level\n risk_level = self._assess_risk(category, text)\n\n return ExtractedClause(\n category=category,\n article_number=section_num,\n title=title,\n full_text=text[:2000], # Limit length\n key_terms=key_terms,\n dollar_amounts=dollar_amounts,\n time_periods=time_periods,\n risk_level=risk_level\n )\n\n def _extract_key_terms(self, category: ClauseCategory, text: str) -> List[str]:\n \"\"\"Extract key terms relevant to clause category.\"\"\"\n key_terms = []\n text_lower = text.lower()\n\n category_terms = {\n ClauseCategory.PAYMENT: [\"net 30\", \"net 45\", \"net 60\", \"retainage\", \"pay when paid\", \"pay if paid\", \"lien waiver\"],\n ClauseCategory.CHANGE_ORDER: [\"written notice\", \"equitable adjustment\", \"constructive change\", \"time extension\"],\n ClauseCategory.SCHEDULE: [\"liquidated damages\", \"substantial completion\", \"final completion\", \"float\", \"concurrent delay\"],\n ClauseCategory.DISPUTE: [\"mediation\", \"arbitration\", \"binding\", \"non-binding\", \"venue\", \"jurisdiction\"],\n ClauseCategory.INSURANCE: [\"additional insured\", \"primary coverage\", \"waiver of subrogation\", \"occurrence form\"],\n ClauseCategory.WARRANTY: [\"one year\", \"two year\", \"workmanship\", \"materials\", \"manufacturer\"],\n ClauseCategory.INDEMNIFICATION: [\"broad form\", \"intermediate form\", \"limited form\", \"negligence\", \"gross negligence\"],\n }\n\n for term in category_terms.get(category, []):\n if term in text_lower:\n key_terms.append(term)\n\n return key_terms\n\n def _assess_risk(self, category: ClauseCategory, text: str) -> RiskLevel:\n \"\"\"Assess risk level of clause.\"\"\"\n text_lower = text.lower()\n\n high_risk_indicators = [\n \"sole discretion\",\n \"waive\",\n \"release\",\n \"indemnify and hold harmless\",\n \"consequential damages\",\n \"no limitation\",\n \"pay if paid\",\n \"broad form indemnification\",\n ]\n\n medium_risk_indicators = [\n \"may require\",\n \"reasonable\",\n \"mutual\",\n \"good faith\",\n ]\n\n high_count = sum(1 for ind in high_risk_indicators if ind in text_lower)\n medium_count = sum(1 for ind in medium_risk_indicators if ind in text_lower)\n\n if high_count >= 2:\n return RiskLevel.HIGH\n elif high_count >= 1 or medium_count >= 2:\n return RiskLevel.MEDIUM\n return RiskLevel.LOW\n\n def generate_summary(self, contract_text: str) -> ContractSummary:\n \"\"\"Generate comprehensive contract summary.\"\"\"\n clauses = self.extract_clauses(contract_text)\n\n # Group clauses by category\n clauses_by_category = {}\n for clause in clauses:\n cat = clause.category.value\n if cat not in clauses_by_category:\n clauses_by_category[cat] = []\n clauses_by_category[cat].append(clause)\n\n # Extract parties\n parties = self._extract_parties(contract_text)\n\n # Extract contract value\n contract_value = self._extract_contract_value(contract_text)\n\n # Risk assessment\n risk_assessment = {}\n for clause in clauses:\n if clause.risk_level == RiskLevel.HIGH:\n risk_assessment[clause.title] = f\"HIGH RISK: Review {clause.category.value} clause carefully\"\n\n # Action items\n action_items = self._generate_action_items(clauses)\n\n return ContractSummary(\n contract_type=self._identify_contract_type(contract_text),\n parties=parties,\n contract_value=contract_value,\n duration_days=0, # Would extract from schedule clause\n start_date=\"\",\n key_dates={},\n clauses_by_category=clauses_by_category,\n risk_assessment=risk_assessment,\n action_items=action_items\n )\n\n def _extract_parties(self, text: str) -> Dict[str, str]:\n \"\"\"Extract contract parties.\"\"\"\n parties = {}\n\n patterns = [\n (r\"Owner[:\\s]+([A-Z][A-Za-z\\s,\\.]+?)(?:\\n|,\\s*(?:a|an))\", \"owner\"),\n (r\"Contractor[:\\s]+([A-Z][A-Za-z\\s,\\.]+?)(?:\\n|,\\s*(?:a|an))\", \"contractor\"),\n (r\"between\\s+([A-Z][A-Za-z\\s,\\.]+?)\\s+\$\\\"Owner\\\"\$\", \"owner\"),\n (r\"and\\s+([A-Z][A-Za-z\\s,\\.]+?)\\s+\$\\\"Contractor\\\"\$\", \"contractor\"),\n ]\n\n for pattern, party_type in patterns:\n match = re.search(pattern, text)\n if match and party_type not in parties:\n parties[party_type] = match.group(1).strip()\n\n return parties\n\n def _extract_contract_value(self, text: str) -> float:\n \"\"\"Extract contract value/sum.\"\"\"\n patterns = [\n r\"contract\\s+(?:sum|price|amount)[:\\s]+\\$?([\\d,]+(?:\\.\\d{2})?)\",\n r\"total\\s+(?:contract|price)[:\\s]+\\$?([\\d,]+(?:\\.\\d{2})?)\",\n r\"\\$?([\\d,]+(?:\\.\\d{2})?)\\s+(?:dollars\\s+)?(?:and\\s+no/100)\",\n ]\n\n for pattern in patterns:\n match = re.search(pattern, text, re.IGNORECASE)\n if match:\n try:\n return float(match.group(1).replace(',', ''))\n except ValueError:\n continue\n\n return 0.0\n\n def _identify_contract_type(self, text: str) -> str:\n \"\"\"Identify contract type.\"\"\"\n text_lower = text.lower()\n\n if \"stipulated sum\" in text_lower or \"lump sum\" in text_lower:\n return \"Stipulated Sum (Lump Sum)\"\n elif \"cost plus\" in text_lower or \"cost of the work\" in text_lower:\n return \"Cost Plus\"\n elif \"guaranteed maximum\" in text_lower or \"gmp\" in text_lower:\n return \"GMP (Guaranteed Maximum Price)\"\n elif \"unit price\" in text_lower:\n return \"Unit Price\"\n elif \"time and materials\" in text_lower or \"t&m\" in text_lower:\n return \"Time and Materials\"\n\n return \"Standard Form\"\n\n def _generate_action_items(self, clauses: List[ExtractedClause]) -> List[str]:\n \"\"\"Generate action items from clause analysis.\"\"\"\n actions = []\n\n for clause in clauses:\n if clause.risk_level == RiskLevel.HIGH:\n actions.append(f\"REVIEW: High-risk {clause.category.value} clause in Article {clause.article_number}\")\n\n if clause.category == ClauseCategory.INSURANCE:\n actions.append(\"Verify insurance requirements meet specified limits\")\n\n if clause.category == ClauseCategory.PAYMENT:\n if \"pay when paid\" in ' '.join(clause.key_terms).lower():\n actions.append(\"ALERT: Pay-when-paid clause detected - assess cash flow risk\")\n if clause.time_periods:\n actions.append(f\"Note payment terms: {', '.join(clause.time_periods)}\")\n\n return list(set(actions))\n\n def generate_report(self, summary: ContractSummary) -> str:\n \"\"\"Generate contract review report.\"\"\"\n lines = [\n \"# Contract Analysis Report\",\n \"\",\n f\"**Contract Type:** {summary.contract_type}\",\n f\"**Contract Value:** ${summary.contract_value:,.2f}\" if summary.contract_value else \"\",\n \"\",\n \"## Parties\",\n \"\"\n ]\n\n for party_type, name in summary.parties.items():\n lines.append(f\"- **{party_type.title()}:** {name}\")\n\n lines.extend([\"\", \"## Key Clauses Summary\", \"\"])\n\n for category, clauses in summary.clauses_by_category.items():\n lines.append(f\"### {category.replace('_', ' ').title()}\")\n for clause in clauses:\n risk_indicator = \"🔴\" if clause.risk_level == RiskLevel.HIGH else \"🟡\" if clause.risk_level == RiskLevel.MEDIUM else \"🟢\"\n lines.append(f\"- {risk_indicator} **Article {clause.article_number}**: {clause.title}\")\n if clause.key_terms:\n lines.append(f\" - Key terms: {', '.join(clause.key_terms)}\")\n lines.append(\"\")\n\n if summary.risk_assessment:\n lines.extend([\"## Risk Assessment\", \"\"])\n for item, note in summary.risk_assessment.items():\n lines.append(f\"- ⚠️ {note}\")\n lines.append(\"\")\n\n if summary.action_items:\n lines.extend([\"## Action Items\", \"\"])\n for item in summary.action_items:\n lines.append(f\"- [ ] {item}\")\n\n return \"\\n\".join(lines)\n```\n\n## Quick Start\n\n```python\n# Initialize extractor\nextractor = ContractClauseExtractor()\n\n# Sample contract text\ncontract_text = \"\"\"\nARTICLE 5 - PAYMENT\n\n5.1 The Contract Sum is Five Million Dollars ($5,000,000.00).\n\n5.2 Progress payments shall be made monthly based on the Schedule of Values.\nRetainage of ten percent (10%) shall be withheld from each payment.\n\n5.3 Final payment shall be made within 30 days of substantial completion.\n\nARTICLE 7 - CHANGES IN THE WORK\n\n7.1 The Owner may order changes in the Work within the general scope of the\nContract. Such changes shall be authorized by written Change Order.\n\n7.2 Contractor shall provide written notice of any claim for additional cost\nor time within 21 days of the event giving rise to such claim.\n\nARTICLE 12 - INDEMNIFICATION\n\n12.1 Contractor shall indemnify and hold harmless the Owner from and against\nall claims arising out of the Contractor's negligence.\n\"\"\"\n\n# Extract clauses\nclauses = extractor.extract_clauses(contract_text)\nfor clause in clauses:\n print(f\"{clause.category.value}: {clause.title} (Risk: {clause.risk_level.value})\")\n\n# Generate summary\nsummary = extractor.generate_summary(contract_text)\nprint(extractor.generate_report(summary))\n```\n\n## Requirements\n\n```bash\npip install (no external dependencies)\n```\n---","attachment_filenames":["claw.json","instructions.md"],"attachments":[{"filename":"claw.json","content":"{\n \"name\": \"contract-clause-extractor\",\n \"version\": \"2.0.0\",\n \"description\": \"Extract and analyze key clauses from construction contracts. Identify payment terms, change order procedures, dispute resolution, warranties, and risk allocation provisions.\",\n \"author\": \"datadrivenconstruction\",\n \"license\": \"MIT\",\n \"permissions\": [\n \"filesystem\"\n ],\n \"entry\": \"instructions.md\",\n \"tags\": [\n \"construction\",\n \"cost-management\"\n ],\n \"models\": [\n \"claude-*\",\n \"gpt-*\"\n ],\n \"minOpenClawVersion\": \"0.8.0\"\n}","content_type":"application/json; charset=utf-8","language":"json","size":525,"content_sha256":"3ccf2d88b1266bb84245a28200fbe071de516ac86c77b005a0899ce04102d547"},{"filename":"instructions.md","content":"You are a construction industry assistant specializing in construction project management.\n\nExtract and analyze key clauses from construction contracts. Identify payment terms, change order procedures, dispute resolution, warranties, and risk allocation provisions.\n\nWhen the user asks to convert or extract data:\n1. Gather the required input data from the user\n2. Process the data using the methods described in SKILL.md\n3. Present results in a clear, structured format\n4. Offer follow-up analysis or export options\n\n## Input Format\n- The user provides project data, file paths, or parameters as described in SKILL.md\n- Accept data in common formats: CSV, Excel, JSON, or direct input\n\n## Output Format\n- Present results in structured tables when applicable\n- Include summary statistics and key findings\n- Offer export to Excel/CSV/JSON when relevant\n\n## Key Reference\n- See SKILL.md for detailed implementation code, classes, and methods\n- Follow the patterns and APIs defined in the skill documentation\n\n## Constraints\n- Only use data provided by the user or referenced in the skill\n- Validate inputs before processing\n- Report errors clearly with suggested fixes\n- Follow construction industry standards and best practices\n","content_type":"text/markdown; charset=utf-8","language":"markdown","size":1227,"content_sha256":"ab4df6a127025b6254a0cf0f3f620a2f24d993b397051e94017fdb66dc70112f"}],"content_json":{"type":"doc","content":[{"type":"heading","attrs":{"level":1},"content":[{"text":"Contract Clause Extractor","type":"text"}]},{"type":"heading","attrs":{"level":2},"content":[{"text":"Overview","type":"text"}]},{"type":"paragraph","content":[{"text":"Extract and analyze key clauses from construction contracts using NLP. Identify critical provisions for payment, changes, disputes, warranties, and risk allocation. Support contract review and compliance tracking.","type":"text"}]},{"type":"heading","attrs":{"level":2},"content":[{"text":"Key Clause Categories","type":"text"}]},{"type":"code_block","attrs":{"wrap":false,"language":""},"content":[{"text":"┌─────────────────────────────────────────────────────────────────┐\n│ CONTRACT CLAUSE CATEGORIES │\n├─────────────────────────────────────────────────────────────────┤\n│ │\n│ Payment Changes Risk │\n│ ─────── ─────── ──── │\n│ 📅 Pay schedule 📝 CO process ⚠️ Indemnification │\n│ 💰 Retainage ⏰ Notice period 🔒 Insurance │\n│ 📋 Requirements 💵 Pricing 🏛️ Liability limits │\n│ │\n│ Schedule Disputes Closeout │\n│ ──────── ──────── ──────── │\n│ 📆 Milestones ⚖️ Resolution ✅ Punch list │\n│ 💸 Liquidated $ 🏛️ Jurisdiction 📄 Warranties │\n│ ⏱️ Extensions 👤 Mediation 🔑 Final payment │\n│ │\n└─────────────────────────────────────────────────────────────────┘","type":"text"}]},{"type":"heading","attrs":{"level":2},"content":[{"text":"Technical Implementation","type":"text"}]},{"type":"code_block","attrs":{"wrap":false,"language":"python"},"content":[{"text":"from dataclasses import dataclass, field\nfrom typing import List, Dict, Optional, Tuple\nfrom enum import Enum\nimport re\n\nclass ClauseCategory(Enum):\n PAYMENT = \"payment\"\n CHANGE_ORDER = \"change_order\"\n SCHEDULE = \"schedule\"\n DISPUTE = \"dispute\"\n INSURANCE = \"insurance\"\n WARRANTY = \"warranty\"\n TERMINATION = \"termination\"\n INDEMNIFICATION = \"indemnification\"\n SAFETY = \"safety\"\n CLOSEOUT = \"closeout\"\n\nclass RiskLevel(Enum):\n HIGH = \"high\"\n MEDIUM = \"medium\"\n LOW = \"low\"\n\n@dataclass\nclass ExtractedClause:\n category: ClauseCategory\n article_number: str\n title: str\n full_text: str\n key_terms: List[str]\n dollar_amounts: List[float]\n time_periods: List[str]\n risk_level: RiskLevel\n notes: str = \"\"\n\n@dataclass\nclass ContractSummary:\n contract_type: str\n parties: Dict[str, str]\n contract_value: float\n duration_days: int\n start_date: str\n key_dates: Dict[str, str]\n clauses_by_category: Dict[str, List[ExtractedClause]]\n risk_assessment: Dict[str, str]\n action_items: List[str]\n\nclass ContractClauseExtractor:\n \"\"\"Extract key clauses from construction contracts.\"\"\"\n\n # Patterns for clause identification\n CLAUSE_PATTERNS = {\n ClauseCategory.PAYMENT: [\n r\"payment\\s+(terms|schedule|application)\",\n r\"progress\\s+payment\",\n r\"retainage|retention\",\n r\"pay\\s+when\\s+paid\",\n r\"pay\\s+if\\s+paid\",\n r\"final\\s+payment\",\n ],\n ClauseCategory.CHANGE_ORDER: [\n r\"change\\s+order\",\n r\"change\\s+directive\",\n r\"modifications?\\s+to\\s+contract\",\n r\"extra\\s+work\",\n r\"changes\\s+in\\s+the\\s+work\",\n ],\n ClauseCategory.SCHEDULE: [\n r\"time\\s+of\\s+completion\",\n r\"liquidated\\s+damages\",\n r\"delay|extension\\s+of\\s+time\",\n r\"substantial\\s+completion\",\n r\"schedule\\s+of\\s+values\",\n r\"milestone\",\n ],\n ClauseCategory.DISPUTE: [\n r\"dispute\\s+resolution\",\n r\"mediation\",\n r\"arbitration\",\n r\"claims?\\s+procedures?\",\n r\"litigation\",\n r\"governing\\s+law\",\n ],\n ClauseCategory.INSURANCE: [\n r\"insurance\\s+requirements?\",\n r\"builder'?s?\\s+risk\",\n r\"general\\s+liability\",\n r\"professional\\s+liability\",\n r\"workers'?\\s+compensation\",\n ],\n ClauseCategory.WARRANTY: [\n r\"warranty|warranties\",\n r\"guarantee\",\n r\"correction\\s+of\\s+work\",\n r\"defect\",\n ],\n ClauseCategory.INDEMNIFICATION: [\n r\"indemnif\",\n r\"hold\\s+harmless\",\n r\"defend\",\n r\"limitation\\s+of\\s+liability\",\n ],\n ClauseCategory.TERMINATION: [\n r\"termination\",\n r\"suspension\\s+of\\s+work\",\n r\"default\",\n r\"for\\s+cause|for\\s+convenience\",\n ],\n }\n\n # Key terms to extract\n KEY_TERM_PATTERNS = {\n \"dollar_amounts\": r'\\$[\\d,]+(?:\\.\\d{2})?|\\d+(?:,\\d{3})*\\s*dollars',\n \"percentages\": r'\\d+(?:\\.\\d+)?%',\n \"time_periods\": r'\\d+\\s*(?:days?|weeks?|months?|years?|business\\s+days?|calendar\\s+days?)',\n \"notice_periods\": r'(?:within|not\\s+(?:less|more)\\s+than)\\s+\\d+\\s*(?:days?|hours?)',\n }\n\n def __init__(self):\n self.extracted_clauses: List[ExtractedClause] = []\n\n def extract_clauses(self, contract_text: str) -> List[ExtractedClause]:\n \"\"\"Extract all identifiable clauses from contract text.\"\"\"\n self.extracted_clauses = []\n\n # Split into articles/sections\n sections = self._split_into_sections(contract_text)\n\n for section_num, section_text in sections.items():\n # Identify clause category\n category = self._identify_category(section_text)\n\n if category:\n clause = self._extract_clause_details(\n category, section_num, section_text\n )\n self.extracted_clauses.append(clause)\n\n return self.extracted_clauses\n\n def _split_into_sections(self, text: str) -> Dict[str, str]:\n \"\"\"Split contract into numbered sections.\"\"\"\n sections = {}\n\n # Pattern for article/section headers\n header_pattern = r'(?:ARTICLE|SECTION|Article|Section)\\s+(\\d+(?:\\.\\d+)?)[:\\.]?\\s*([A-Z][A-Za-z\\s]+)?'\n\n parts = re.split(header_pattern, text)\n\n current_num = \"0\"\n current_text = \"\"\n\n for i, part in enumerate(parts):\n if re.match(r'^\\d+(?:\\.\\d+)?

Important: agents should read /llm.txt, /llms.txt, or /.well-known/skills.json to discover the public Skillopedia API.