Risk Assessment

Field	Value
Document ID	ASCEND-GOV-007
Version	1.0.0
Last Updated	December 19, 2025
Author	Ascend Engineering Team
Publisher	OW-KAI Technologies Inc.
Classification	Enterprise Client Documentation
Compliance	SOC 2 CC6.1/CC6.2, PCI-DSS 7.1/8.3, HIPAA 164.312, NIST 800-53 AC-2/SI-4

Reading Time: 12 minutes | Skill Level: Advanced

Overview

ASCEND's risk assessment engine calculates dynamic risk scores (0-100) for every AI agent action. The score determines whether actions are auto-approved, require human approval, or are escalated.

Risk Calculation Formula

┌─────────────────────────────────────────────────────────────────────────────────────┐
│                          RISK SCORE CALCULATION                                      │
├─────────────────────────────────────────────────────────────────────────────────────┤
│                                                                                      │
│   Risk Score = (Environment × 35%) + (Data Sensitivity × 33%)                       │
│              + (Action Type × 25%) + (Operational Context × 7%)                     │
│                                                                                      │
│   ┌──────────────────────────────────────────────────────────────────────────────┐  │
│   │                                                                               │  │
│   │     ENVIRONMENT          DATA SENSITIVITY      ACTION TYPE        CONTEXT    │  │
│   │     Weight: 35%          Weight: 33%           Weight: 25%        Weight: 7% │  │
│   │                                                                               │  │
│   │     production: 35       high_pii: 30          delete: 25         peak: +10  │  │
│   │     staging:    20       medium_pii: 20        write: 20          night: +5  │  │
│   │     development: 5       low_pii: 10           read: 10           normal: 0  │  │
│   │                          none: 0               describe: 5                   │  │
│   │                                                                               │  │
│   │     × Resource Multiplier (0.8 - 1.2)                                        │  │
│   │     • database: 1.2x                                                          │  │
│   │     • secrets: 1.2x                                                           │  │
│   │     • storage: 1.1x                                                           │  │
│   │     • compute: 0.9x                                                           │  │
│   │                                                                               │  │
│   └──────────────────────────────────────────────────────────────────────────────┘  │
│                                                                                      │
│   Example: DELETE on PRODUCTION DATABASE with PII                                   │
│   = (35 × 0.35) + (30 × 0.33) + (25 × 0.25) + (0 × 0.07) × 1.2                      │
│   = (12.25 + 9.9 + 6.25 + 0) × 1.2                                                  │
│   = 28.4 × 1.2 = 34.08 → Risk Score: 34                                             │
│                                                                                      │
└─────────────────────────────────────────────────────────────────────────────────────┘

Configuration API

Get Active Configuration

curl "https://pilot.owkai.app/api/risk-scoring/config" \
  -H "Authorization: Bearer owkai_..."

Response:

{
  "id": 1,
  "config_version": "1.2.0",
  "algorithm_version": "2.0.0",
  "environment_weights": {
    "production": 35,
    "staging": 20,
    "development": 5
  },
  "action_weights": {
    "delete": 25,
    "write": 20,
    "read": 10,
    "describe": 5,
    "list": 8
  },
  "resource_multipliers": {
    "rds": 1.2,
    "dynamodb": 1.2,
    "s3": 1.1,
    "lambda": 0.9,
    "ec2": 1.0,
    "iam": 1.2,
    "secretsmanager": 1.2,
    "kms": 1.2
  },
  "pii_weights": {
    "high_sensitivity": 30,
    "medium_sensitivity": 20,
    "low_sensitivity": 10,
    "none": 0
  },
  "component_percentages": {
    "environment": 35,
    "data_sensitivity": 33,
    "action_type": 25,
    "operational_context": 7
  },
  "is_active": true,
  "created_at": "2025-12-01T10:00:00Z",
  "activated_at": "2025-12-01T10:30:00Z"
}

Create New Configuration

curl -X POST "https://pilot.owkai.app/api/risk-scoring/config" \
  -H "Authorization: Bearer owkai_..." \
  -H "Content-Type: application/json" \
  -d '{
    "config_version": "1.3.0",
    "algorithm_version": "2.0.0",
    "description": "Updated weights for financial compliance",
    "environment_weights": {
      "production": 40,
      "staging": 25,
      "development": 5
    },
    "action_weights": {
      "delete": 30,
      "write": 25,
      "read": 15,
      "describe": 5,
      "list": 10
    },
    "resource_multipliers": {
      "rds": 1.3,
      "dynamodb": 1.2,
      "s3": 1.1,
      "lambda": 0.8,
      "ec2": 1.0,
      "iam": 1.3,
      "secretsmanager": 1.3,
      "kms": 1.3
    },
    "pii_weights": {
      "high_sensitivity": 35,
      "medium_sensitivity": 25,
      "low_sensitivity": 10,
      "none": 0
    },
    "component_percentages": {
      "environment": 35,
      "data_sensitivity": 35,
      "action_type": 23,
      "operational_context": 7
    }
  }'

Validate Configuration

curl -X POST "https://pilot.owkai.app/api/risk-scoring/config/validate" \
  -H "Authorization: Bearer owkai_..." \
  -H "Content-Type: application/json" \
  -d '{
    "config_version": "1.3.0",
    "component_percentages": {
      "environment": 35,
      "data_sensitivity": 35,
      "action_type": 25,
      "operational_context": 10
    }
  }'

Response:

{
  "valid": false,
  "errors": [
    "Component percentages must sum to 100 (currently 105)"
  ],
  "warnings": [
    "High production weight (40) may cause excessive approvals"
  ]
}

Activate Configuration

curl -X PUT "https://pilot.owkai.app/api/risk-scoring/config/5/activate" \
  -H "Authorization: Bearer owkai_..."

Response:

{
  "id": 5,
  "config_version": "1.3.0",
  "is_active": true,
  "activated_at": "2025-12-15T10:30:00Z",
  "activated_by": "admin@company.com",
  "previous_config_id": 4
}

Weight Configuration

Environment Weights (0-100)

Environment	Default	Description
`production`	35	Live customer data
`staging`	20	Pre-production testing
`development`	5	Developer environments

Action Weights (0-100)

Action	Default	Description
`delete`	25	Data destruction
`write`	20	Data modification
`read`	10	Data access
`list`	8	Resource enumeration
`describe`	5	Metadata access

Resource Multipliers (0.8-1.2)

Resource	Multiplier	Rationale
`rds`	1.2	Relational databases
`dynamodb`	1.2	NoSQL databases
`secretsmanager`	1.2	Credentials
`kms`	1.2	Encryption keys
`iam`	1.2	Identity management
`s3`	1.1	Object storage
`ec2`	1.0	Compute (neutral)
`lambda`	0.9	Functions (lower risk)

PII Weights (0-100)

Classification	Default	Examples
`high_sensitivity`	30	SSN, health records, financial data
`medium_sensitivity`	20	Email, phone, address
`low_sensitivity`	10	Name, general PII
`none`	0	Non-PII data

Risk Score Interpretation

Risk Levels

Score Range	Level	Typical Workflow
0-30	Low	Auto-approve
30-60	Medium	Single approval
60-80	High	Multi-approval
80-100	Critical	Security escalation

Example Calculations

Low Risk Action

# Read from S3 in development
action = {
    "action_type": "read",
    "environment": "development",
    "resource": "s3",
    "data_classification": "none"
}

# Calculation:
# Environment: 5 × 0.35 = 1.75
# Data: 0 × 0.33 = 0
# Action: 10 × 0.25 = 2.5
# Context: 0 × 0.07 = 0
# Subtotal: 4.25
# × Multiplier: 4.25 × 1.1 = 4.68

# Risk Score: 5 (LOW)

High Risk Action

# Delete from production database with PII
action = {
    "action_type": "delete",
    "environment": "production",
    "resource": "rds",
    "data_classification": "high_sensitivity"
}

# Calculation:
# Environment: 35 × 0.35 = 12.25
# Data: 30 × 0.33 = 9.9
# Action: 25 × 0.25 = 6.25
# Context: 0 × 0.07 = 0
# Subtotal: 28.4
# × Multiplier: 28.4 × 1.2 = 34.08

# Risk Score: 34 (MEDIUM)
# Note: Would be higher with peak hours context

CVSS Integration

CVSS 3.1 Mapping

ASCEND can map risk scores to CVSS severity levels:

CVSS Severity	CVSS Score	ASCEND Risk
None	0.0	0-10
Low	0.1-3.9	10-30
Medium	4.0-6.9	30-60
High	7.0-8.9	60-80
Critical	9.0-10.0	80-100

Enable CVSS Mapping

{
  "cvss_mapping_enabled": true,
  "cvss_thresholds": {
    "low": 10,
    "medium": 30,
    "high": 60,
    "critical": 80
  }
}

Configuration History

List Configuration History

curl "https://pilot.owkai.app/api/risk-scoring/config/history?limit=10" \
  -H "Authorization: Bearer owkai_..."

Response:

{
  "configs": [
    {
      "id": 5,
      "config_version": "1.3.0",
      "is_active": true,
      "created_at": "2025-12-15T10:00:00Z",
      "created_by": "admin@company.com",
      "activated_at": "2025-12-15T10:30:00Z"
    },
    {
      "id": 4,
      "config_version": "1.2.0",
      "is_active": false,
      "created_at": "2025-12-01T10:00:00Z",
      "created_by": "admin@company.com",
      "activated_at": "2025-12-01T10:30:00Z"
    }
  ]
}

Rollback to Default

curl -X POST "https://pilot.owkai.app/api/risk-scoring/config/rollback-to-default" \
  -H "Authorization: Bearer owkai_..."

Factory Default

New organizations receive a factory default configuration:

# Source: routes/risk_scoring_config_routes.py:47
factory_default = {
    "config_version": "1.0.0-default",
    "algorithm_version": "2.0.0",
    "environment_weights": {
        "production": 35,
        "staging": 20,
        "development": 5
    },
    "action_weights": {
        "delete": 25,
        "write": 20,
        "read": 10,
        "describe": 5,
        "list": 8
    },
    "resource_multipliers": {
        "rds": 1.2,
        "dynamodb": 1.2,
        "s3": 1.1,
        "lambda": 0.9,
        "ec2": 1.0,
        "iam": 1.2,
        "secretsmanager": 1.2,
        "kms": 1.2
    },
    "pii_weights": {
        "high_sensitivity": 30,
        "medium_sensitivity": 20,
        "low_sensitivity": 10,
        "none": 0
    },
    "component_percentages": {
        "environment": 35,
        "data_sensitivity": 33,
        "action_type": 25,
        "operational_context": 7
    }
}

SDK Integration

Get Risk Breakdown

from ascend import AscendClient

client = AscendClient(api_key="owkai_...")

result = client.evaluate_action(
    action_type="database.delete",
    resource="customers",
    parameters={"table": "users"},
    context={
        "environment": "production",
        "data_classification": "pii"
    }
)

# Access risk breakdown
print(f"Risk Score: {result.risk_score}")
print(f"Risk Level: {result.risk_level}")
print(f"Factors:")
for factor, contribution in result.risk_factors.items():
    print(f"  {factor}: {contribution}")

Custom Risk Indicators

result = client.evaluate_action(
    action_type="financial.transfer",
    resource="wire_transfer",
    parameters={"amount": 50000},
    risk_indicators={
        "risk_level": "high",
        "data_classification": "financial",
        "requires_approval": True
    }
)

Audit Trail

All configuration changes are logged:

{
  "event_type": "CONFIG_CHANGE",
  "actor_id": "admin@company.com",
  "resource_type": "RISK_CONFIG",
  "action": "ACTIVATE",
  "event_data": {
    "config_id": 5,
    "config_version": "1.3.0",
    "previous_config_id": 4
  },
  "risk_level": "HIGH",
  "compliance_tags": ["SOX", "CONFIG_MANAGEMENT", "CRITICAL_CHANGE"]
}

Best Practices

1. Start with Defaults

# Use factory defaults initially
# Monitor and tune based on your patterns

2. Validate Before Activating

# Always validate configuration
validation = client.validate_risk_config(new_config)

if validation.errors:
    print("Cannot activate - fix errors first")
else:
    client.activate_risk_config(new_config.id)

3. Version Configurations

# Use semantic versioning
{
    "config_version": "1.3.0"  # MAJOR.MINOR.PATCH
}
# MAJOR: Breaking changes
# MINOR: New weights/resources
# PATCH: Threshold adjustments

4. Document Changes

{
    "description": "JIRA-1234: Increased production weight per security audit"
}

5. Test in Staging

# Create config without activating
POST /api/risk-scoring/config

# Test with dry-run
POST /api/risk-scoring/config/validate

# Review in staging before production activation

Next Steps

Policies — Configure governance policies
Audit Logging — Track changes
Compliance — Meet regulatory requirements

Document Version: 1.0.0 | Last Updated: December 2025

Overview​

Risk Calculation Formula​

Configuration API​

Get Active Configuration​

Create New Configuration​

Validate Configuration​

Activate Configuration​

Weight Configuration​

Environment Weights (0-100)​

Action Weights (0-100)​

Resource Multipliers (0.8-1.2)​

PII Weights (0-100)​

Risk Score Interpretation​

Risk Levels​

Example Calculations​

Low Risk Action​

High Risk Action​

CVSS Integration​

CVSS 3.1 Mapping​

Enable CVSS Mapping​

Configuration History​

List Configuration History​

Rollback to Default​

Factory Default​

SDK Integration​

Get Risk Breakdown​

Custom Risk Indicators​

Audit Trail​

Best Practices​

1. Start with Defaults​

2. Validate Before Activating​

3. Version Configurations​

4. Document Changes​

5. Test in Staging​

Next Steps​

Overview

Risk Calculation Formula

Configuration API

Get Active Configuration

Create New Configuration

Validate Configuration

Activate Configuration

Weight Configuration

Environment Weights (0-100)

Action Weights (0-100)

Resource Multipliers (0.8-1.2)

PII Weights (0-100)

Risk Score Interpretation

Risk Levels

Example Calculations

Low Risk Action

High Risk Action

CVSS Integration

CVSS 3.1 Mapping

Enable CVSS Mapping

Configuration History

List Configuration History

Rollback to Default

Factory Default

SDK Integration

Get Risk Breakdown

Custom Risk Indicators

Audit Trail

Best Practices

1. Start with Defaults

2. Validate Before Activating

3. Version Configurations

4. Document Changes

5. Test in Staging

Next Steps