Security Review Construction

Security Review Skill for Construction Systems

This skill ensures all construction software systems follow security best practices, protecting sensitive project data, financial information, and business intelligence.

When to Activate

• Building ERP/BIM system integrations
• Creating construction dashboards
• Handling cost/financial data
• Building document management systems
• Creating APIs for field data collection
• Integrating with external platforms (Procore, PlanGrid, etc.)
• Working with subcontractor/vendor data
• Processing payment applications

Construction-Specific Security Concerns

1. Financial Data Protection

# CRITICAL: Construction financial data security

# ❌ NEVER Do This
project_budget = 15000000  # Hardcoded in source
margin_percentage = 0.18   # Business-sensitive info in code

# ✅ ALWAYS Do This
import os
from cryptography.fernet import Fernet

# Load from secure configuration
project_config = load_secure_config(os.environ['PROJECT_CONFIG_PATH'])

# Encrypt sensitive data at rest
def encrypt_financial_data(data: dict) -> bytes:
    key = os.environ.get('ENCRYPTION_KEY')
    f = Fernet(key)
    return f.encrypt(json.dumps(data).encode())

Financial Data Checklist

• Cost estimates encrypted at rest
• Margin/markup data not exposed in logs
• Payment information tokenized
• Historical pricing protected from competitors
• Bid amounts secured until opening

2. BIM/CAD Data Security

# BIM data often contains proprietary design information

# ❌ NEVER store BIM directly in public cloud without encryption
s3.upload_file('model.ifc', bucket='public-bucket')

# ✅ ALWAYS encrypt and control access
def upload_bim_secure(file_path: str, project_id: str):
    # Encrypt file
    encrypted_path = encrypt_file(file_path)

    # Generate pre-signed URL with expiration
    presigned_url = s3.generate_presigned_url(
        'get_object',
        Params={
            'Bucket': 'secure-bim-bucket',
            'Key': f'{project_id}/{os.path.basename(file_path)}'
        },
        ExpiresIn=3600  # 1 hour expiration
    )

    # Log access
    audit_log.info(f"BIM access granted: {project_id}")

    return presigned_url

BIM/CAD Checklist

• IFC/RVT files encrypted at rest
• Access logged for audit trail
• Time-limited download links
• Version control with access tracking
• No design data in error messages

3. Subcontractor/Vendor Data

# Subcontractor data includes business-sensitive information

class SubcontractorDataHandler:
    """Secure handling of subcontractor data"""

    # Fields that require encryption
    SENSITIVE_FIELDS = [
        'insurance_policy_number',
        'bank_account',
        'tax_id',
        'bonding_capacity',
        'historical_pricing'
    ]

    def store_subcontractor(self, data: dict) -> str:
        # Encrypt sensitive fields
        for field in self.SENSITIVE_FIELDS:
            if field in data:
                data[field] = self.encrypt(data[field])

        # Store with audit trail
        sub_id = self.db.insert(data)
        self.audit.log(f"Subcontractor created: {sub_id}")

        return sub_id

    def get_subcontractor(self, sub_id: str, requester_id: str) -> dict:
        # Check authorization
        if not self.can_access(requester_id, sub_id):
            raise PermissionError("Unauthorized access to subcontractor data")

        # Log access
        self.audit.log(f"Subcontractor accessed: {sub_id} by {requester_id}")

        # Return with decrypted sensitive fields (only to authorized users)
        return self.decrypt_sensitive_fields(self.db.get(sub_id))

Vendor Data Checklist

• Insurance/bonding information encrypted
• Bank details protected (PCI compliance)
• Tax IDs masked in UI (show last 4 digits only)
• Pricing history access-controlled
• Certificate expiration notifications secure

4. Field Data Collection Security

# Mobile/field data collection must be secure

from datetime import datetime, timedelta
import hashlib

class FieldDataCollector:
    """Secure field data collection"""

    def validate_photo_submission(self, photo_data: dict) -> bool:
        # Verify GPS timestamp is recent (within 24 hours)
        photo_time = datetime.fromisoformat(photo_data['timestamp'])
        if datetime.now() - photo_time > timedelta(hours=24):
            raise ValueError("Photo timestamp too old - possible replay attack")

        # Verify file hash matches
        file_hash = hashlib.sha256(photo_data['content']).hexdigest()
        if file_hash != photo_data['declared_hash']:
            raise ValueError("File integrity check failed")

        # Validate GPS coordinates are within project boundary
        if not self.is_within_project_bounds(
            photo_data['lat'],
            photo_data['lon'],
            photo_data['project_id']
        ):
            self.audit.warn(f"Photo from outside project bounds: {photo_data}")

        return True

    def submit_daily_report(self, report: dict, user_id: str) -> str:
        # Verify user is assigned to project
        if not self.is_assigned_to_project(user_id, report['project_id']):
            raise PermissionError("User not assigned to this project")

        # Sign report with user credentials
        report['signature'] = self.sign_report(report, user_id)
        report['submitted_at'] = datetime.now().isoformat()

        return self.db.insert(report)

Field Data Checklist

• GPS data validated for reasonableness
• Photo timestamps verified
• File integrity checks (hashing)
• User authentication for submissions
• Offline data sync secured

5. CWICR Database Security

# CWICR contains proprietary cost data

class CWICRAccessControl:
    """Access control for CWICR database"""

    TIERS = {
        'basic': ['public_rates', 'standard_descriptions'],
        'professional': ['regional_rates', 'productivity_factors'],
        'enterprise': ['custom_rates', 'historical_data', 'analytics']
    }

    def search(self, query: str, user_id: str) -> list:
        # Get user tier
        tier = self.get_user_tier(user_id)

        # Limit results based on tier
        allowed_fields = self.TIERS[tier]

        # Execute search with field restrictions
        results = self.vector_search(
            query=query,
            fields=allowed_fields,
            limit=self.get_tier_limit(tier)
        )

        # Log search for analytics
        self.audit.log(f"CWICR search: {user_id}, query='{query[:50]}...'")

        return results

    def export_data(self, user_id: str, format: str) -> bytes:
        # Enterprise only
        if self.get_user_tier(user_id) != 'enterprise':
            raise PermissionError("Export requires enterprise tier")

        # Watermark exported data
        data = self.get_exportable_data(user_id)
        watermarked = self.add_watermark(data, user_id)

        return watermarked

CWICR Checklist

• Tiered access control implemented
• API rate limiting per user/tier
• Data exports watermarked
• Bulk download restrictions
• Competitor access monitoring

6. Integration Security (Procore, PlanGrid, etc.)

# Secure OAuth integration with construction platforms

class ConstructionPlatformIntegration:
    """Secure integration with external platforms"""

    def __init__(self, platform: str):
        self.platform = platform
        # Load credentials from secure vault
        self.credentials = self.vault.get(f'{platform}_oauth')

    def authenticate(self) -> str:
        # Use OAuth 2.0 with PKCE
        code_verifier = secrets.token_urlsafe(32)
        code_challenge = base64.urlsafe_b64encode(
            hashlib.sha256(code_verifier.encode()).digest()
        ).decode().rstrip('=')

        # Never store tokens in code or logs
        token = self.oauth_flow(code_verifier, code_challenge)

        # Store token securely
        self.secure_token_store.set(
            key=f'{self.platform}_token',
            value=token,
            ttl=token['expires_in']
        )

        return token

    def sync_data(self, project_id: str) -> dict:
        # Validate project access before sync
        if not self.has_project_access(project_id):
            raise PermissionError(f"No access to project {project_id}")

        # Rate limit syncs
        self.rate_limiter.check(f'sync_{self.platform}')

        # Sync with retry and error handling
        try:
            data = self.api_client.get_project_data(project_id)
            self.validate_incoming_data(data)
            return data
        except APIError as e:
            # Log error without sensitive details
            self.logger.error(f"Sync failed for {project_id}: {type(e).__name__}")
            raise

Integration Checklist

• OAuth 2.0 with PKCE implemented
• Tokens stored in secure vault (not env vars)
• Token refresh automated
• API rate limiting respected
• Webhook signatures verified
• Data validation on incoming data

7. Document Management Security

# Construction documents often contain confidential information

class SecureDocumentManager:
    """Secure document handling for construction"""

    # Document classification levels
    CLASSIFICATIONS = {
        'public': [],
        'internal': ['daily_reports', 'schedules'],
        'confidential': ['contracts', 'bids', 'financials'],
        'restricted': ['legal', 'hr', 'insurance']
    }

    def upload_document(self, file: bytes, metadata: dict, user_id: str) -> str:
        # Scan for malware
        if not self.malware_scan(file):
            raise SecurityError("Malware detected in uploaded file")

        # Classify document
        classification = self.classify_document(metadata)

        # Check user can upload to this classification
        if not self.can_upload(user_id, classification):
            raise PermissionError(f"Cannot upload {classification} documents")

        # Encrypt based on classification
        if classification in ['confidential', 'restricted']:
            file = self.encrypt(file)

        # Store with audit trail
        doc_id = self.storage.put(file, metadata)
        self.audit.log(f"Document uploaded: {doc_id} by {user_id}")

        return doc_id

    def download_document(self, doc_id: str, user_id: str) -> bytes:
        # Check access
        doc = self.storage.get_metadata(doc_id)
        if not self.can_access(user_id, doc['classification']):
            raise PermissionError("Access denied")

        # Log download
        self.audit.log(f"Document downloaded: {doc_id} by {user_id}")

        # Return decrypted content
        return self.decrypt(self.storage.get(doc_id))

Document Checklist

• Malware scanning on upload
• Document classification system
• Role-based access control
• Download audit logging
• Encryption for sensitive documents
• Retention policies enforced

Pre-Deployment Security Checklist for Construction Systems

Data Protection

• Financial data encrypted at rest and in transit
• BIM/CAD files protected with access control
• Subcontractor PII secured (GDPR/CCPA compliant)
• CWICR data access tiered appropriately
• Backup encryption enabled

Authentication & Authorization

• Multi-factor authentication for admin users
• Role-based access control implemented
• Session management secure (timeout, single device)
• API keys rotated regularly
• OAuth integrations use PKCE

Audit & Compliance

• All data access logged
• Logs tamper-proof (append-only)
• Retention policies documented
• Data export capabilities for audits
• Compliance reports automated

Integration Security

• All external APIs authenticated
• Webhook signatures verified
• Data validation on all inputs
• Rate limiting implemented
• Error messages sanitized

Field Data Security

• Mobile apps use certificate pinning
• Offline data encrypted
• GPS/timestamp validation
• Photo integrity verification
• Secure sync protocols

Resources

• OWASP Top 10
• NIST Cybersecurity Framework
• CIS Controls for Construction
• ISO 27001 for Construction

Remember: Construction data includes financial, legal, and competitive information. A breach can result in lost bids, legal liability, and reputational damage. Security is not optional.