Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add jdmorag97-rgb/DDC_Skills_for_AI_Agents_in_Construction --skill cwicr-data-loadergit clone --depth 1 https://github.com/jdmorag97-rgb/DDC_Skills_for_AI_Agents_in_ConstructionWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/jdmorag97-rgb/ddc_skills_for_ai_agents_in_construction/cwicr-data-loader)<a href="https://agentmods.dev/skills/jdmorag97-rgb/ddc_skills_for_ai_agents_in_construction/cwicr-data-loader"><img src="https://agentmods.dev/badge/skills/jdmorag97-rgb/ddc_skills_for_ai_agents_in_construction/cwicr-data-loader/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/jdmorag97-rgb/ddc_skills_for_ai_agents_in_construction/cwicr-data-loader"><img src="https://agentmods.dev/badge/skills/jdmorag97-rgb/ddc_skills_for_ai_agents_in_construction/cwicr-data-loader.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00041 | $0.03516 |
| Opus 5 | $0.00020 | $0.01758 |
| Sonnet 5 | $0.00008 | $0.00703 |
| Haiku 4.5 | $0.00004 | $0.00352 |
Grade A, and why
cwicr-data-loader scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
This is a copy
100% identical to cwicr-data-loader — 2 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 471 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CWICR Data Loader
Business Case
Problem Statement
DDC CWICR database is distributed in multiple formats:
- Apache Parquet (optimized for analytics)
- Excel workbooks (human-readable)
- CSV files (universal exchange)
- Qdrant snapshots (vector search)
Applications need unified data access regardless of source format.
Solution
Universal data loader supporting all CWICR formats with automatic schema detection, validation, and pandas DataFrame conversion.
Business Value
- Format agnostic - Load from any CWICR distribution
- Validated data - Automatic schema validation
- Memory efficient - Lazy loading for large datasets
- Type-safe - Proper data types preserved
Technical Implementation
Prerequisites
pip install pandas pyarrow openpyxl qdrant-client
Python Implementation
import pandas as pd
import pyarrow.parquet as pq
from pathlib import Path
from typing import Optional, Dict, Any, List, Union
from dataclasses import dataclass, field
from enum import Enum
import json
class CWICRFormat(Enum):
"""Supported CWICR data formats."""
PARQUET = "parquet"
EXCEL = "excel"
CSV = "csv"
QDRANT = "qdrant"
JSON = "json"
class CWICRLanguage(Enum):
"""Supported languages in CWICR database."""
ARABIC = "ar"
CHINESE = "zh"
GERMAN = "de"
ENGLISH = "en"
SPANISH = "es"
FRENCH = "fr"
HINDI = "hi"
PORTUGUESE = "pt"
RUSSIAN = "ru"
@dataclass
class CWICRSchema:
"""CWICR database schema definition."""
# Core fields
work_item_code: str = "work_item_code"
description: str = "description"
unit: str = "unit"
category: str = "category"
# Cost fields
unit_price: str = "unit_price"
labor_cost: str = "labor_cost"
material_cost: str = "material_cost"
equipment_cost: str = "equipment_cost"
overhead_cost: str = "overhead_cost"
# Norm fields
labor_norm: str = "labor_norm"
material_norm: str = "material_norm"
equipment_norm: str = "equipment_norm"
# Metadata
language: str = "language"
region: str = "region"
currency: str = "currency"
last_updated: str = "last_updated"
# Optional embedding
embedding: str = "embedding"
@dataclass
class CWICRWorkItem:
"""Represents a single work item from CWICR database."""
work_item_code: str
description: str
unit: str
category: str
unit_price: float = 0.0
labor_cost: float = 0.0
material_cost: float = 0.0
equipment_cost: float = 0.0
overhead_cost: float = 0.0
labor_norm: float = 0.0
labor_unit: str = "h"
resources: List[Dict[str, Any]] = field(default_factory=list)
language: str = "en"
region: str = ""
currency: str = "USD"
@dataclass
class CWICRResource:
"""Represents a resource (material, labor, equipment)."""
resource_code: str
description: str
unit: str
unit_price: float
resource_type: str # 'labor', 'material', 'equipment'
category: str = ""
class CWICRDataLoader:
"""Universal loader for CWICR database formats."""
REQUIRED_COLUMNS = ['work_item_code', 'description', 'unit']
NUMERIC_COLUMNS = ['unit_price', 'labor_cost', 'material_cost',
'equipment_cost', 'labor_norm']
def __init__(self):
self.schema = CWICRSchema()
self._cache: Dict[str, pd.DataFrame] = {}
def load(self, source: str,
format: Optional[CWICRFormat] = None,
language: Optional[CWICRLanguage] = None,
use_cache: bool = True) -> pd.DataFrame:
"""Load CWICR data from any supported source."""
cache_key = f"{source}_{language}"
if use_cache and cache_key in self._cache:
return self._cache[cache_key]
# Auto-detect format if not specified
if format is None:
format = self._detect_format(source)
# Load based on format
if format == CWICRFormat.PARQUET:
df = self._load_parquet(source)
elif format == CWICRFormat.EXCEL:
df = self._load_excel(source)
elif format == CWICRFormat.CSV:
df = self._load_csv(source)
elif format == CWICRFormat.JSON:
df = self._load_json(source)
else:
raise ValueError(f"Unsupported format: {format}")
# Validate and normalize
df = self._validate_schema(df)
df = self._normalize_types(df)
# Filter by language if specified
if language and 'language' in df.columns:
df = df[df['language'] == language.value]
# Cache result
if use_cache:
self._cache[cache_key] = df
return df
def _detect_format(self, source: str) -> CWICRFormat:
"""Auto-detect data format from source."""
path = Path(source)
if path.suffix.lower() == '.parquet':
return CWICRFormat.PARQUET
elif path.suffix.lower() in ['.xlsx', '.xls']:
return CWICRFormat.EXCEL
elif path.suffix.lower() == '.csv':
return CWICRFormat.CSV
elif path.suffix.lower() == '.json':
return CWICRFormat.JSON
else:
raise ValueError(f"Cannot detect format: {source}")
def _load_parquet(self, source: str) -> pd.DataFrame:
"""Load from Parquet file."""
return pd.read_parquet(source)
def _load_excel(self, source: str,
sheet_name: str = "WorkItems") -> pd.DataFrame:
"""Load from Excel workbook."""
try:
return pd.read_excel(source, sheet_name=sheet_name)
except:
# Try first sheet if named sheet doesn't exist
return pd.read_excel(source, sheet_name=0)
def _load_csv(self, source: str) -> pd.DataFrame:
"""Load from CSV file."""
# Try different encodings
for encoding in ['utf-8', 'latin-1', 'cp1252']:
try:
return pd.read_csv(source, encoding=encoding)
except UnicodeDecodeError:
continue
raise ValueError(f"Cannot read CSV with any encoding: {source}")
def _load_json(self, source: str) -> pd.DataFrame:
"""Load from JSON file."""
with open(source, 'r', encoding='utf-8') as f:
data = json.load(f)
if isinstance(data, list):
return pd.DataFrame(data)
elif isinstance(data, dict) and 'items' in data:
return pd.DataFrame(data['items'])
else:
return pd.DataFrame([data])
def _validate_schema(self, df: pd.DataFrame) -> pd.DataFrame:
"""Validate DataFrame against CWICR schema."""
# Check required columns
missing = set(self.REQUIRED_COLUMNS) - set(df.columns)
if missing:
raise ValueError(f"Missing required columns: {missing}")
return df
def _normalize_types(self, df: pd.DataFrame) -> pd.DataFrame:
"""Normalize column types."""
for col in self.NUMERIC_COLUMNS:
if col in df.columns:
df[col] = pd.to_numeric(df[col], errors='coerce').fillna(0)
# Ensure string columns
for col in ['work_item_code', 'description', 'unit', 'category']:
if col in df.columns:
df[col] = df[col].astype(str)
return df
def load_resources(self, source: str,
format: Optional[CWICRFormat] = None) -> pd.DataFrame:
"""Load resources separately."""
if format is None:
format = self._detect_format(source)
if format == CWICRFormat.EXCEL:
try:
return pd.read_excel(source, sheet_name="Resources")
except:
return pd.DataFrame()
else:
return self.load(source, format)
def get_work_item(self, df: pd.DataFrame,
code: str) -> Optional[CWICRWorkItem]:
"""Get single work item by code."""
item = df[df['work_item_code'] == code]
if item.empty:
return None
row = item.iloc[0]
return CWICRWorkItem(
work_item_code=row['work_item_code'],
description=row.get('description', ''),
unit=row.get('unit', ''),
category=row.get('category', ''),
unit_price=row.get('unit_price', 0),
labor_cost=row.get('labor_cost', 0),
material_cost=row.get('material_cost', 0),
equipment_cost=row.get('equipment_cost', 0),
labor_norm=row.get('labor_norm', 0),
language=row.get('language', 'en'),
region=row.get('region', ''),
currency=row.get('currency', 'USD')
)
def get_categories(self, df: pd.DataFrame) -> List[str]:
"""Get unique categories."""
if 'category' not in df.columns:
return []
return df['category'].dropna().unique().tolist()
def filter_by_category(self, df: pd.DataFrame,
category: str) -> pd.DataFrame:
"""Filter work items by category."""
return df[df['category'] == category]
def search_by_description(self, df: pd.DataFrame,
keyword: str,
case_sensitive: bool = False) -> pd.DataFrame:
"""Simple keyword search in descriptions."""
if case_sensitive:
return df[df['description'].str.contains(keyword, na=False)]
return df[df['description'].str.contains(keyword, case=False, na=False)]
def get_statistics(self, df: pd.DataFrame) -> Dict[str, Any]:
"""Get database statistics."""
stats = {
'total_work_items': len(df),
'categories': df['category'].nunique() if 'category' in df.columns else 0,
'languages': df['language'].unique().tolist() if 'language' in df.columns else ['en']
}
if 'unit_price' in df.columns:
stats['price_range'] = {
'min': df['unit_price'].min(),
'max': df['unit_price'].max(),
'mean': df['unit_price'].mean()
}
return stats
def export(self, df: pd.DataFrame,
output_path: str,
format: CWICRFormat = CWICRFormat.PARQUET):
"""Export DataFrame to file."""
if format == CWICRFormat.PARQUET:
df.to_parquet(output_path, index=False)
elif format == CWICRFormat.EXCEL:
df.to_excel(output_path, index=False)
elif format == CWICRFormat.CSV:
df.to_csv(output_path, index=False)
elif format == CWICRFormat.JSON:
df.to_json(output_path, orient='records', indent=2)
class CWICRBatchLoader:
"""Load multiple CWICR files and merge."""
def __init__(self):
self.loader = CWICRDataLoader()
def load_multiple(self, sources: List[str]) -> pd.DataFrame:
"""Load and merge multiple CWICR files."""
dfs = []
for source in sources:
try:
df = self.loader.load(source)
dfs.append(df)
except Exception as e:
print(f"Warning: Failed to load {source}: {e}")
if not dfs:
return pd.DataFrame()
return pd.concat(dfs, ignore_index=True)
def load_all_languages(self, base_path: str) -> pd.DataFrame:
"""Load all language variants from directory."""
path = Path(base_path)
dfs = []
for lang in CWICRLanguage:
# Try various naming patterns
patterns = [
f"cwicr_{lang.value}.*",
f"ddc_cwicr_{lang.value}.*",
f"*_{lang.value}.*"
]
for pattern in patterns:
files = list(path.glob(pattern))
for file in files:
try:
df = self.loader.load(str(file), language=lang)
dfs.append(df)
except Exception as e:
continue
if not dfs:
return pd.DataFrame()
return pd.concat(dfs, ignore_index=True)
# Convenience functions
def load_cwicr(source: str, language: str = None) -> pd.DataFrame:
"""Quick load CWICR data."""
loader = CWICRDataLoader()
lang = CWICRLanguage(language) if language else None
return loader.load(source, language=lang)
def get_cwicr_statistics(source: str) -> Dict[str, Any]:
"""Get statistics from CWICR source."""
loader = CWICRDataLoader()
df = loader.load(source)
return loader.get_statistics(df)
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 471 lines · 41 tokens per session scan A cd99f507a4be
cwicr-data-loader is a skill published in the GitHub repository jdmorag97-rgb/DDC_Skills_for_AI_Agents_in_Construction (2 stars, last pushed 6mo ago), licensed MIT. It adds 41 tokens to every session and 3,516 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to cwicr-data-loader, differing in 2 lines, and is treated as a copy.
Other skills, from other repositories
create-pr
Creates a GitHub PR with a Linear-ticket-prefixed title and a decision-led, narrative description for Prisma 8. Use when the user wants to create a pull request, open a PR, or submit changes for review.
schema-exploration
Lists tables, describes columns and data types, identifies foreign key relationships, and maps entity relationships in a database. Use when the user asks about database schema, table structure, column types, what tables exist, ERD, foreign keys, or how entities relate.
ha-data-stores
Map of Hope Agent's local data stores and safe read-only query workflow. Use when the user asks where Hope Agent stores data, wants to inspect sessions/messages/memory/logs/background jobs/knowledge indexes/settings, asks the model to query local app data, or debugging requires checking persisted state. Trigger…
nornicdb-cypher-queries
Pick fast, predictable Cypher query shapes in NornicDB — point lookups, batch retrieval, pagination, search, traversal, batched UNWIND/MERGE writes, cleanup, multi-tenant isolation. Use when writing or reviewing Cypher whose latency or throughput matters; maps user intent to the executor's hot-path query templates.
supabase
Supabase / PostgREST Row-Level-Security playbook — pull the anon (or leaked servicerole) key out of the frontend JS, map tables from the auto-generated OpenAPI spec, test anonymous RLS READ disclosures (PII/secret leaks), and anonymous RLS WRITE abuse (insert/update/delete — e.g. forging…
volcengine-rds-postgresql
A tool for operating PostgreSQL databases hosted by Volcano Engine's managed database service. PostgreSQL is a relational database used to store structured application data.