cwicr-value-engineering — скилл для Claude Code: обзор и запуск
Скилл cwicr-value-engineering анализирует строительные сметы по базе CWICR, находит более дешёвые альтернативы материалам и фиксирует экономию.
MIT 116 установок ★ 434 обновлён 2026-10-19
npx skills add datadrivenconstruction/DDC_Skills_for_AI_Agents_in_Construction --skill cwicr-value-engineering -a claude-codeЧто делает cwicr-value-engineering
Скилл выполняет стоимостной анализ строительных смет с помощью датасета CWICR. Логика сопоставляет позиции внутри похожих категорий, сравнивая затраты на материалы, труд и спецтехнику.
Внутри работает модуль CWICRValueEngineering на базе pandas и numpy. Он находит подходящие замены по префиксам категорий и собирает карточки VEProposal. В каждом предложении фиксируются расчётная экономия, возможные риски, сдвиг сроков и влияние на качество.
Для запуска обязателен python3. Без загруженного датафрейма с колонкой work_item_code функции поиска возвращают пустой список.
Какие задачи решает
- подобрать дешёвые строительные альтернативы по коду работы
- рассчитать разбивку стоимости по материалам, труду и технике
- оформить предложение VEProposal с оценкой рисков и качества
- отслеживать статусы согласования инженерных предложений
Примеры запросов агенту
- Найди альтернативы для позиции с кодом work_item_code через cwicr-value-engineering
- Создай предложение VEProposal по замене материала с расчётом экономии
Как установить cwicr-value-engineering
1Через Skills CLI (Vercel)
npx skills add datadrivenconstruction/DDC_Skills_for_AI_Agents_in_Construction --skill cwicr-value-engineering -a claude-code2Вручную: положить файлы
~/.claude/skills/cwicr-value-engineering/SKILL.mdПапка скилла целиком; для одного проекта — .claude/skills/cwicr-value-engineering/
1Через Skills CLI (Vercel)
npx skills add datadrivenconstruction/DDC_Skills_for_AI_Agents_in_Construction --skill cwicr-value-engineering -a codex2Вручную: положить файлы
.agents/skills/cwicr-value-engineering/SKILL.mdДля всех проектов — ~/.agents/skills/
1Через Skills CLI (Vercel)
npx skills add datadrivenconstruction/DDC_Skills_for_AI_Agents_in_Construction --skill cwicr-value-engineering -a cursor2Вручную: положить файлы
.cursor/skills/cwicr-value-engineering/SKILL.mdCursor читает и .agents/skills/, и .claude/skills/
1GitHub CLI
gh skill install datadrivenconstruction/DDC_Skills_for_AI_Agents_in_Construction 1_DDC_Toolkit/CWICR-Database/cwicr-value-engineering2Вручную: положить файлы
.github/skills/cwicr-value-engineering/SKILL.md1Через Skills CLI (Vercel)
npx skills add datadrivenconstruction/DDC_Skills_for_AI_Agents_in_Construction --skill cwicr-value-engineering -a gemini-cli2Командой платформы
gemini skills install https://github.com/datadrivenconstruction/DDC_Skills_for_AI_Agents_in_Construction --path 1_DDC_Toolkit/CWICR-Database/cwicr-value-engineeringВручную: положить файлы
.devin/skills/cwicr-value-engineering/SKILL.mdСтарый путь .windsurf/skills/ тоже читается
Через Skills CLI (Vercel)
npx skills add datadrivenconstruction/DDC_Skills_for_AI_Agents_in_Construction --skill cwicr-value-engineering -a opencodeClaude.ai
Упакуйте папку в zip и загрузите в Settings → Capabilities → Skills
По шагам, с командами для каждой платформы: Как установить скилл в Claude Code: пять способов
Вопросы о cwicr-value-engineering
Как установить cwicr-value-engineering в Claude Code?
Выполните npx skills add datadrivenconstruction/DDC_Skills_for_AI_Agents_in_Construction --skill cwicr-value-engineering -a claude-code. Файл навыка запишется в ~/.claude/skills/cwicr-value-engineering/SKILL.md.
Что делает cwicr-value-engineering?
Скилл автоматизирует поиск более дешёвых проектных решений по базе данных CWICR и ведёт учёт предложений по оптимизации бюджета.
Какие системные требования нужны для работы?
Требуется интерпретатор python3, а также библиотеки pandas и numpy для обработки таблиц с расценками.
Исходный текст
SKILL.md · MIT
---
name: "cwicr-value-engineering"
description: "Perform value engineering analysis using CWICR data. Identify cost-saving alternatives while maintaining function and quality."
homepage: "https://datadrivenconstruction.io"
metadata: {"openclaw": {"emoji": "🗄️", "os": ["darwin", "linux", "win32"], "homepage": "https://datadrivenconstruction.io", "requires": {"bins": ["python3"]}}}
---
# CWICR Value Engineering
## Business Case
### Problem Statement
Projects often exceed budget:
- Where can costs be reduced?
- What alternatives exist?
- How to maintain quality?
- Document VE decisions
### Solution
Systematic value engineering using CWICR data to identify cost-effective alternatives, analyze trade-offs, and document decisions.
### Business Value
- **Cost savings** - Identify reduction opportunities
- **Quality maintenance** - Function-based analysis
- **Documentation** - VE proposal records
- **Client value** - Optimize value for cost
## Technical Implementation
```python
import pandas as pd
import numpy as np
from typing import Dict, Any, List, Optional, Tuple
from dataclasses import dataclass, field
from datetime import date
from enum import Enum
class VECategory(Enum):
"""Value engineering categories."""
MATERIAL = "material"
METHOD = "method"
DESIGN = "design"
SPECIFICATION = "specification"
SYSTEM = "system"
class VEStatus(Enum):
"""VE proposal status."""
PROPOSED = "proposed"
UNDER_REVIEW = "under_review"
ACCEPTED = "accepted"
REJECTED = "rejected"
IMPLEMENTED = "implemented"
@dataclass
class VEProposal:
"""Value engineering proposal."""
proposal_id: str
title: str
category: VECategory
description: str
original_item: str
proposed_item: str
original_cost: float
proposed_cost: float
savings: float
savings_percent: float
function_impact: str
quality_impact: str
schedule_impact: int
risk_assessment: str
status: VEStatus
@dataclass
class VEAnalysis:
"""Complete VE analysis."""
project_name: str
total_original_cost: float
total_proposed_cost: float
total_savings: float
savings_percent: float
proposals: List[VEProposal]
accepted_savings: float
pending_savings: float
class CWICRValueEngineering:
"""Value engineering analysis using CWICR data."""
def __init__(self, cwicr_data: pd.DataFrame):
self.cost_data = cwicr_data
self._index_data()
self._proposals: Dict[str, VEProposal] = {}
def _index_data(self):
"""Index cost data."""
if 'work_item_code' in self.cost_data.columns:
self._code_index = self.cost_data.set_index('work_item_code')
else:
self._code_index = None
def get_item_cost(self, code: str, quantity: float = 1) -> Tuple[float, Dict[str, float]]:
"""Get item cost breakdown."""
if self._code_index is None or code not in self._code_index.index:
return (0, {})
item = self._code_index.loc[code]
labor = float(item.get('labor_cost', 0) or 0) * quantity
material = float(item.get('material_cost', 0) or 0) * quantity
equipment = float(item.get('equipment_cost', 0) or 0) * quantity
return (labor + material + equipment, {
'labor': labor,
'material': material,
'equipment': equipment
})
def find_alternatives(self,
work_item_code: str,
quantity: float,
max_cost_increase: float = 0) -> List[Dict[str, Any]]:
"""Find alternative work items that could replace original."""
original_cost, _ = self.get_item_cost(work_item_code, quantity)
if self._code_index is None:
return []
# Get original item category
if work_item_code in self._code_index.index:
original = self._code_index.loc[work_item_code]
category = str(original.get('category', '')).lower()
else:
return []
alternatives = []
for code, row in self._code_index.iterrows():
if code == work_item_code:
continue
# Match by category prefix or similar category
item_category = str(row.get('category', '')).lower()
if category[:4] in item_category or item_category[:4] in category:
alt_cost, breakdown = self.get_item_cost(code, quantity)
if alt_cost <= original_cost * (1 + max_cost_increase):
savings = original_cost - alt_cost
alternatives.append({
'code': code,
'description': str(row.get('description', code)),
'cost': round(alt_cost, 2),
'savings': round(savings, 2),
'savings_pct': round(savings / original_cost * 100, 1) if original_cost > 0 else 0,
'breakdown': breakdown
})
# Sort by savings
return sorted(alternatives, key=lambda x: x['savings'], reverse=True)[:10]
def create_proposal(self,
proposal_id: str,
title: str,
category: VECategory,
description: str,
original_item: str,
proposed_item: str,
quantity: float,
function_impact: str = "Equivalent",
quality_impact: str = "Equivalent",
schedule_impact: int = 0,
risk_assessment: str = "Low") -> VEProposal:
"""Create VE proposal."""
original_cost, _ = self.get_item_cost(original_item, quantity)
proposed_cost, _ = self.get_item_cost(proposed_item, quantity)
savings = original_cost - proposed_cost
savings_pct = (savings / original_cost * 100) if original_cost > 0 else 0
proposal = VEProposal(
proposal_id=proposal_id,
title=title,
category=category,
description=description,
original_item=original_item,
proposed_item=proposed_item,
original_cost=round(original_cost, 2),
proposed_cost=round(proposed_cost, 2),
savings=round(savings, 2),
savings_percent=round(savings_pct, 1),
function_impact=function_impact,
quality_impact=quality_impact,
schedule_impact=schedule_impact,
risk_assessment=risk_assessment,
status=VEStatus.PROPOSED
)
self._proposals[proposal_id] = proposal
return proposal
def update_status(self, proposal_id: str, status: VEStatus):
"""Update proposal status."""
if proposal_id in self._proposals:
self._proposals[proposal_id].status = status
def identify_high_cost_items(self,
items: List[Dict[str, Any]],
top_n: int = 20,
min_percentage: float = 2.0) -> List[Dict[str, Any]]:
"""Identify high-cost items for VE focus."""
item_costs = []
total_cost = 0
for item in items:
code = item.get('work_item_code', item.get('code'))
qty = item.get('quantity', 0)
cost, breakdown = self.get_item_cost(code, qty)
item_costs.append({
'code': code,
'quantity': qty,
'cost': cost,
'breakdown': breakdown
})
total_cost += cost
# Add percentage and sort
for item in item_costs:
item['percentage'] = round(item['cost'] / total_cost * 100, 2) if total_cost > 0 else 0
# Filter and sort
significant = [i for i in item_costs if i['percentage'] >= min_percentage]
significant.sort(key=lambda x: x['cost'], reverse=True)
return significant[:top_n]
def analyze_material_alternatives(self,
material_type: str,
quantity: float) -> Dict[str, Any]:
"""Analyze alternative materials by type."""
if self._code_index is None:
return {}
matches = []
for code, row in self._code_index.iterrows():
desc = str(row.get('description', '')).lower()
if material_type.lower() in desc:
cost, breakdown = self.get_item_cost(code, quantity)
matches.append({
'code': code,
'description': str(row.get('description', code)),
'cost': cost,
'material_cost': breakdown.get('material', 0),
'unit': str(row.get('unit', 'unit'))
})
if not matches:
return {}
matches.sort(key=lambda x: x['cost'])
cheapest = matches[0]
most_expensive = matches[-1]
return {
'material_type': material_type,
'quantity': quantity,
'options_found': len(matches),
'cheapest': cheapest,
'most_expensive': most_expensive,
'potential_savings': round(most_expensive['cost'] - cheapest['cost'], 2),
'all_options': matches
}
def generate_ve_analysis(self, project_name: str) -> VEAnalysis:
"""Generate complete VE analysis."""
proposals = list(self._proposals.values())
total_original = sum(p.original_cost for p in proposals)
total_proposed = sum(p.proposed_cost for p in proposals)
total_savings = sum(p.savings for p in proposals)
accepted_savings = sum(
p.savings for p in proposals
if p.status in [VEStatus.ACCEPTED, VEStatus.IMPLEMENTED]
)
pending_savings = sum(
p.savings for p in proposals
if p.status in [VEStatus.PROPOSED, VEStatus.UNDER_REVIEW]
)
return VEAnalysis(
project_name=project_name,
total_original_cost=round(total_original, 2),
total_proposed_cost=round(total_proposed, 2),
total_savings=round(total_savings, 2),
savings_percent=round(total_savings / total_original * 100, 1) if total_original > 0 else 0,
proposals=proposals,
accepted_savings=round(accepted_savings, 2),
pending_savings=round(pending_savings, 2)
)
def export_ve_report(self,
analysis: VEAnalysis,
output_path: str) -> str:
"""Export VE analysis to Excel."""
with pd.ExcelWriter(output_path, engine='openpyxl') as writer:
# Summary
summary_df = pd.DataFrame([{
'Project': analysis.project_name,
'Total Original Cost': analysis.total_original_cost,
'Total Proposed Cost': analysis.total_proposed_cost,
'Total Savings': analysis.total_savings,
'Savings %': analysis.savings_percent,
'Accepted Savings': analysis.accepted_savings,
'Pending Savings': analysis.pending_savings
}])
summary_df.to_excel(writer, sheet_name='Summary', index=False)
# Proposals
proposals_df = pd.DataFrame([
{
'ID': p.proposal_id,
'Title': p.title,
'Category': p.category.value,
'Original Item': p.original_item,
'Proposed Item': p.proposed_item,
'Original Cost': p.original_cost,
'Proposed Cost': p.proposed_cost,
'Savings': p.savings,
'Savings %': p.savings_percent,
'Function Impact': p.function_impact,
'Quality Impact': p.quality_impact,
'Schedule Days': p.schedule_impact,
'Risk': p.risk_assessment,
'Status': p.status.value
}
for p in analysis.proposals
])
proposals_df.to_excel(writer, sheet_name='Proposals', index=False)
return output_path
```
## Quick Start
```python
# Load CWICR data
cwicr = pd.read_parquet("TR_workitems_costs_resources_DDC_CWICR.parquet")
# Initialize VE analyzer
ve = CWICRValueEngineering(cwicr)
# Find alternatives for expensive item
alternatives = ve.find_alternatives(
work_item_code="CONC-HIGH-001",
quantity=100
)
for alt in alternatives[:3]:
print(f"{alt['code']}: ${alt['savings']:,.2f} savings ({alt['savings_pct']}%)")
```
## Common Use Cases
### 1. Identify VE Opportunities
```python
items = [
{'work_item_code': 'CONC-001', 'quantity': 200},
{'work_item_code': 'STRL-002', 'quantity': 50}
]
high_cost = ve.identify_high_cost_items(items, top_n=10, min_percentage=5.0)
for item in high_cost:
print(f"{item['code']}: ${item['cost']:,.2f} ({item['percentage']}%)")
```
### 2. Create VE Proposal
```python
proposal = ve.create_proposal(
proposal_id="VE-001",
title="Substitute concrete grade",
category=VECategory.MATERIAL,
description="Use C25 instead of C30 for non-structural elements",
original_item="CONC-C30-001",
proposed_item="CONC-C25-001",
quantity=150,
function_impact="Equivalent for intended use",
quality_impact="Meets specification",
risk_assessment="Low"
)
print(f"Potential Savings: ${proposal.savings:,.2f}")
```
### 3. Generate VE Report
```python
analysis = ve.generate_ve_analysis("Building Project")
ve.export_ve_report(analysis, "ve_analysis.xlsx")
```
## Resources
- **GitHub**: [OpenConstructionEstimate-DDC-CWICR](https://github.com/datadrivenconstruction/OpenConstructionEstimate-DDC-CWICR)
- **DDC Book**: Chapter 3.2 - Value Engineering
© datadrivenconstruction · MIT · datadrivenconstruction/DDC_Skills_for_AI_Agents_in_Construction