Construction safety incident reporting and analysis. Capture incidents, conduct investigations, track corrective actions, and analyze trends for prevention.
npx skills add https://github.com/datadrivenconstruction/DDC_Skills_for_AI_Agents_in_Construction --skill incident-reporting
Comprehensive incident reporting system for construction safety. Capture near-misses, injuries, and property damage. Conduct root cause analysis and track corrective actions to prevent recurrence.
> "Near-miss reporting prevents 90% of future serious incidents" — DDC Community
△
/│\ Fatality (1)
/ │ \
/ │ \ Serious Injury (10)
/ │ \
/ │ \ Minor Injury (30)
/ │ \
/ │ \ Near Miss (300)
/ │ \
/ │ \ Unsafe Acts (3000)
──────────┴──────────
from dataclasses import dataclass, field
from typing import List, Dict, Optional
from enum import Enum
from datetime import datetime, timedelta
import json
class IncidentType(Enum):
NEAR_MISS = "near_miss"
FIRST_AID = "first_aid"
MEDICAL_TREATMENT = "medical_treatment"
LOST_TIME = "lost_time"
FATALITY = "fatality"
PROPERTY_DAMAGE = "property_damage"
ENVIRONMENTAL = "environmental"
class IncidentCategory(Enum):
FALL = "fall"
STRUCK_BY = "struck_by"
CAUGHT_IN = "caught_in"
ELECTROCUTION = "electrocution"
VEHICLE = "vehicle"
MATERIAL_HANDLING = "material_handling"
TOOL_EQUIPMENT = "tool_equipment"
SLIP_TRIP = "slip_trip"
FIRE = "fire"
CHEMICAL = "chemical"
OTHER = "other"
class InvestigationStatus(Enum):
REPORTED = "reported"
UNDER_INVESTIGATION = "under_investigation"
ROOT_CAUSE_IDENTIFIED = "root_cause_identified"
CORRECTIVE_ACTIONS_ASSIGNED = "corrective_actions_assigned"
IN_REMEDIATION = "in_remediation"
CLOSED = "closed"
@dataclass
class Person:
name: str
company: str
role: str
contact: str
years_experience: int = 0
@dataclass
class CorrectiveAction:
id: str
description: str
assigned_to: str
due_date: datetime
status: str = "open"
completed_date: Optional[datetime] = None
verification_notes: str = ""
@dataclass
class Incident:
id: str
incident_type: IncidentType
category: IncidentCategory
date_time: datetime
location: str
project_id: str
project_name: str
# Description
description: str
immediate_actions: str
# People involved
injured_person: Optional[Person] = None
witnesses: List[Person] = field(default_factory=list)
reported_by: str = ""
# Investigation
status: InvestigationStatus = InvestigationStatus.REPORTED
root_causes: List[str] = field(default_factory=list)
contributing_factors: List[str] = field(default_factory=list)
corrective_actions: List[CorrectiveAction] = field(default_factory=list)
# Documentation
photos: List[str] = field(default_factory=list)
weather_conditions: str = ""
equipment_involved: List[str] = field(default_factory=list)
# Metrics
days_lost: int = 0
property_damage_cost: float = 0.0
osha_recordable: bool = False
class IncidentManager:
"""Manage construction incident reporting and investigation."""
# 5 Whys root cause categories
ROOT_CAUSE_CATEGORIES = [
"Training/Competency",
"Procedures/Work Instructions",
"Equipment/Tools",
"Supervision",
"Communication",
"Housekeeping",
"PPE",
"Work Environment",
"Physical/Mental State",
"Management System"
]
def __init__(self):
self.incidents: Dict[str, Incident] = {}
self.corrective_actions: Dict[str, CorrectiveAction] = {}
def report_incident(self, incident_type: IncidentType,
category: IncidentCategory,
date_time: datetime,
location: str,
project_id: str,
project_name: str,
description: str,
immediate_actions: str,
reported_by: str,
injured_person: Dict = None) -> Incident:
"""Report new incident."""
incident_id = f"INC-{datetime.now().strftime('%Y%m%d%H%M%S')}"
injured = None
if injured_person:
injured = Person(**injured_person)
incident = Incident(
id=incident_id,
incident_type=incident_type,
category=category,
date_time=date_time,
location=location,
project_id=project_id,
project_name=project_name,
description=description,
immediate_actions=immediate_actions,
reported_by=reported_by,
injured_person=injured
)
# Auto-flag OSHA recordable
if incident_type in [IncidentType.MEDICAL_TREATMENT,
IncidentType.LOST_TIME,
IncidentType.FATALITY]:
incident.osha_recordable = True
self.incidents[incident_id] = incident
return incident
def add_witness(self, incident_id: str, witness: Dict) -> Incident:
"""Add witness to incident."""
if incident_id not in self.incidents:
raise ValueError(f"Incident {incident_id} not found")
self.incidents[incident_id].witnesses.append(Person(**witness))
return self.incidents[incident_id]
def conduct_investigation(self, incident_id: str,
root_causes: List[str],
contributing_factors: List[str]) -> Incident:
"""Record investigation findings."""
if incident_id not in self.incidents:
raise ValueError(f"Incident {incident_id} not found")
incident = self.incidents[incident_id]
incident.root_causes = root_causes
incident.contributing_factors = contributing_factors
incident.status = InvestigationStatus.ROOT_CAUSE_IDENTIFIED
return incident
def five_whys_analysis(self, incident_id: str, whys: List[str]) -> Dict:
"""Conduct 5 Whys analysis."""
if incident_id not in self.incidents:
raise ValueError(f"Incident {incident_id} not found")
analysis = {
"incident_id": incident_id,
"analysis_date": datetime.now().isoformat(),
"whys": []
}
for i, why in enumerate(whys):
analysis["whys"].append({
"level": i + 1,
"question": f"Why #{i+1}?",
"answer": why
})
# The last "why" is typically the root cause
if whys:
self.incidents[incident_id].root_causes.append(whys[-1])
return analysis
def assign_corrective_action(self, incident_id: str,
description: str,
assigned_to: str,
due_days: int = 7) -> CorrectiveAction:
"""Assign corrective action."""
if incident_id not in self.incidents:
raise ValueError(f"Incident {incident_id} not found")
action_id = f"CA-{datetime.now().strftime('%Y%m%d%H%M%S')}"
action = CorrectiveAction(
id=action_id,
description=description,
assigned_to=assigned_to,
due_date=datetime.now() + timedelta(days=due_days)
)
self.incidents[incident_id].corrective_actions.append(action)
self.incidents[incident_id].status = InvestigationStatus.CORRECTIVE_ACTIONS_ASSIGNED
self.corrective_actions[action_id] = action
return action
def complete_corrective_action(self, action_id: str,
verification_notes: str) -> CorrectiveAction:
"""Mark corrective action complete."""
if action_id not in self.corrective_actions:
raise ValueError(f"Corrective action {action_id} not found")
action = self.corrective_actions[action_id]
action.status = "completed"
action.completed_date = datetime.now()
action.verification_notes = verification_notes
return action
def get_incident_metrics(self, project_id: str = None,
start_date: datetime = None,
end_date: datetime = None) -> Dict:
"""Calculate incident metrics."""
incidents = list(self.incidents.values())
if project_id:
incidents = [i for i in incidents if i.project_id == project_id]
if start_date:
incidents = [i for i in incidents if i.date_time >= start_date]
if end_date:
incidents = [i for i in incidents if i.date_time <= end_date]
# Calculate metrics
total = len(incidents)
near_misses = len([i for i in incidents if i.incident_type == IncidentType.NEAR_MISS])
first_aid = len([i for i in incidents if i.incident_type == IncidentType.FIRST_AID])
recordables = len([i for i in incidents if i.osha_recordable])
lost_time = len([i for i in incidents if i.incident_type == IncidentType.LOST_TIME])
total_days_lost = sum(i.days_lost for i in incidents)
# Category breakdown
by_category = {}
for cat in IncidentCategory:
count = len([i for i in incidents if i.category == cat])
if count > 0:
by_category[cat.value] = count
return {
"total_incidents": total,
"near_misses": near_misses,
"first_aid_cases": first_aid,
"osha_recordables": recordables,
"lost_time_incidents": lost_time,
"total_days_lost": total_days_lost,
"by_category": by_category,
"near_miss_ratio": near_misses / recordables if recordables else 0
}
def calculate_trir(self, hours_worked: int, project_id: str = None) -> float:
"""Calculate Total Recordable Incident Rate."""
incidents = list(self.incidents.values())
if project_id:
incidents = [i for i in incidents if i.project_id == project_id]
recordables = len([i for i in incidents if i.osha_recordable])
if hours_worked == 0:
return 0
# TRIR = (Recordables × 200,000) / Hours Worked
return (recordables * 200000) / hours_worked
def calculate_dart(self, hours_worked: int, project_id: str = None) -> float:
"""Calculate Days Away, Restricted, or Transferred rate."""
incidents = list(self.incidents.values())
if project_id:
incidents = [i for i in incidents if i.project_id == project_id]
dart_cases = len([i for i in incidents
if i.incident_type in [IncidentType.LOST_TIME]])
if hours_worked == 0:
return 0
return (dart_cases * 200000) / hours_worked
def get_trend_analysis(self, months: int = 6) -> List[Dict]:
"""Analyze incident trends over time."""
trends = []
now = datetime.now()
for i in range(months):
month_start = datetime(now.year, now.month - i, 1) if now.month > i else datetime(now.year - 1, 12 - (i - now.month), 1)
month_end = month_start.replace(day=28) + timedelta(days=4)
month_end = month_end - timedelta(days=month_end.day)
month_incidents = [inc for inc in self.incidents.values()
if month_start <= inc.date_time <= month_end]
trends.append({
"month": month_start.strftime("%Y-%m"),
"total": len(month_incidents),
"near_misses": len([i for i in month_incidents if i.incident_type == IncidentType.NEAR_MISS]),
"recordables": len([i for i in month_incidents if i.osha_recordable])
})
return list(reversed(trends))
def generate_incident_report(self, incident_id: str) -> str:
"""Generate detailed incident report."""
if incident_id not in self.incidents:
return "Incident not found"
inc = self.incidents[incident_id]
lines = [
f"# Incident Report",
f"",
f"**Incident ID:** {inc.id}",
f"**Type:** {inc.incident_type.value}",
f"**Category:** {inc.category.value}",
f"**Date/Time:** {inc.date_time.strftime('%Y-%m-%d %H:%M')}",
f"**Location:** {inc.location}",
f"**Project:** {inc.project_name}",
f"**Status:** {inc.status.value}",
f"**OSHA Recordable:** {'Yes' if inc.osha_recordable else 'No'}",
f"",
f"## Description",
f"{inc.description}",
f"",
f"## Immediate Actions Taken",
f"{inc.immediate_actions}",
f"",
]
if inc.injured_person:
lines.extend([
f"## Injured Person",
f"- Name: {inc.injured_person.name}",
f"- Company: {inc.injured_person.company}",
f"- Role: {inc.injured_person.role}",
f"- Experience: {inc.injured_person.years_experience} years",
f""
])
if inc.root_causes:
lines.extend([
f"## Root Causes",
*[f"- {rc}" for rc in inc.root_causes],
f""
])
if inc.corrective_actions:
lines.extend([
f"## Corrective Actions",
f"| Action | Assigned To | Due | Status |",
f"|--------|-------------|-----|--------|"
])
for ca in inc.corrective_actions:
lines.append(f"| {ca.description} | {ca.assigned_to} | {ca.due_date.strftime('%Y-%m-%d')} | {ca.status} |")
return "\n".join(lines)
# Initialize manager
manager = IncidentManager()
# Report near-miss
incident = manager.report_incident(
incident_type=IncidentType.NEAR_MISS,
category=IncidentCategory.STRUCK_BY,
date_time=datetime.now(),
location="Level 5, Grid C-4",
project_id="PRJ-001",
project_name="Office Tower",
description="Unsecured tool fell from scaffold, landed 2 feet from worker",
immediate_actions="Area cordoned off, toolbox talk conducted",
reported_by="Site Foreman"
)
# 5 Whys analysis
analysis = manager.five_whys_analysis(incident.id, [
"Tool fell from scaffold",
"Tool was not secured",
"No tool lanyard was used",
"Worker not trained on tool tethering",
"Tool tethering training not included in orientation"
])
# Assign corrective actions
manager.assign_corrective_action(
incident.id,
"Update orientation to include tool tethering training",
"Safety Manager",
due_days=14
)
manager.assign_corrective_action(
incident.id,
"Provide tool lanyards to all workers at height",
"Procurement",
due_days=7
)
# Get metrics
metrics = manager.get_incident_metrics()
print(f"Total Incidents: {metrics['total_incidents']}")
print(f"Near Miss Ratio: {metrics['near_miss_ratio']:.1f}")
# Generate report
print(manager.generate_incident_report(incident.id))
pip install (no external dependencies)
Comprehensive spreadsheet creation, editing, and analysis with support for formulas, formatting, data analysis, and visualization. When Claude needs to work with spreadsheets (.xlsx, .xlsm, .csv, .tsv, etc) for: (1) Creating new spreadsheets with formulas and formatting, (2) Reading or analyzing data, (3) Modify existing spreadsheets while preserving formulas, (4) Data analysis and visualization in spreadsheets, or (5) Recalculating formulas
Use this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like \"the xlsx in my downloads\") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved.
Picks random winners from lists, spreadsheets, or Google Sheets for giveaways, raffles, and contests. Ensures fair, unbiased selection with transparency.
Query openFDA API for drugs, devices, adverse events, recalls, regulatory submissions (510k, PMA), substance identification (UNII), for FDA regulatory data analysis and safety research.
MATLAB and GNU Octave numerical computing for matrix operations, data analysis, visualization, and scientific computing. Use when writing MATLAB/Octave scripts for linear algebra, signal processing, image processing, differential equations, optimization, statistics, or creating scientific visualizations. Also use when the user needs help with MATLAB syntax, functions, or wants to convert between MATLAB and Python code. Scripts can be executed with MATLAB or the open-source GNU Octave interpreter.
UMAP dimensionality reduction. Fast nonlinear manifold learning for 2D/3D visualization, clustering preprocessing (HDBSCAN), supervised/parametric UMAP, for high-dimensional data.
Creating interactive data visualisations using d3.js. This skill should be used when creating custom charts, graphs, network diagrams, geographic visualisations, or any complex SVG-based data visualisation that requires fine-grained control over visual elements, transitions, or interactions. Use this for bespoke visualisations beyond standard charting libraries, whether in React, Vue, Svelte, vanilla JavaScript, or any other environment.
Access AlphaFold 200M+ AI-predicted protein structures. Retrieve structures by UniProt ID, download PDB/mmCIF files, analyze confidence metrics (pLDDT, PAE), for drug discovery and structural biology.
Take datadrivenconstruction/incident-reporting from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.
The instructions reference pip.
Without those the skill loads but fails at the first command.