Security

Security API Reference

Complete API reference for NeuronEdge security features: guardrail rules, presets, security events, and red team scanning.

Authentication

All endpoints require an Authorization header with a valid API key:

Authorization: Bearer <api-key>

Security endpoints require Professional tier or above.

Guardrail Rules

GET/api/security/rules

List all guardrail rules. Supports filtering by category, action, and pagination.

Request Body

bash

curl -X GET "https://api.neuronedge.ai/api/security/rules?category=prompt_injection&limit=20" \
  -H "Authorization: Bearer <api-key>"

# Response
{
  "rules": [
    {
      "id": "rule_abc123",
      "name": "Block SQL Injection Attempts",
      "category": "prompt_injection",
      "pattern": "(?i)(union|select|insert|drop|delete)\\s+(from|into|table)",
      "action": "block",
      "description": "Detects common SQL injection patterns",
      "priority": 100,
      "enabled": true,
      "created_at": "2026-03-01T10:00:00Z"
    }
  ],
  "total": 42,
  "limit": 20,
  "offset": 0
}

POST/api/security/rules

Create a custom guardrail rule with specified pattern, category, and action.

Request Body

bash

curl -X POST "https://api.neuronedge.ai/api/security/rules" \
  -H "Authorization: Bearer <api-key>" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Block Credit Card Numbers",
    "category": "pii_leakage",
    "pattern": "\\b\\d{4}[\\s-]?\\d{4}[\\s-]?\\d{4}[\\s-]?\\d{4}\\b",
    "action": "redact",
    "description": "Detects and redacts credit card numbers",
    "priority": 90
  }'

# Response
{
  "id": "rule_xyz789",
  "name": "Block Credit Card Numbers",
  "category": "pii_leakage",
  "pattern": "\\b\\d{4}[\\s-]?\\d{4}[\\s-]?\\d{4}[\\s-]?\\d{4}\\b",
  "action": "redact",
  "description": "Detects and redacts credit card numbers",
  "priority": 90,
  "enabled": true,
  "created_at": "2026-03-10T14:30:00Z"
}

GET/api/security/rules/:id

Retrieve details for a specific guardrail rule, including match statistics.

Request Body

bash

curl -X GET "https://api.neuronedge.ai/api/security/rules/rule_abc123" \
  -H "Authorization: Bearer <api-key>"

# Response
{
  "id": "rule_abc123",
  "name": "Block SQL Injection Attempts",
  "category": "prompt_injection",
  "pattern": "(?i)(union|select|insert|drop|delete)\\s+(from|into|table)",
  "action": "block",
  "description": "Detects common SQL injection patterns",
  "priority": 100,
  "enabled": true,
  "created_at": "2026-03-01T10:00:00Z",
  "updated_at": "2026-03-05T08:15:00Z",
  "match_count": 1247
}

PATCH/api/security/rules/:id

Update an existing rule. Supports partial updates of any field.

Request Body

bash

curl -X PATCH "https://api.neuronedge.ai/api/security/rules/rule_abc123" \
  -H "Authorization: Bearer <api-key>" \
  -H "Content-Type: application/json" \
  -d '{
    "enabled": false,
    "priority": 80
  }'

# Response
{
  "id": "rule_abc123",
  "name": "Block SQL Injection Attempts",
  "enabled": false,
  "priority": 80,
  "updated_at": "2026-03-10T14:35:00Z"
}

DELETE/api/security/rules/:id

Delete a guardrail rule permanently.

Request Body

bash

curl -X DELETE "https://api.neuronedge.ai/api/security/rules/rule_abc123" \
  -H "Authorization: Bearer <api-key>"

# Response
{
  "deleted": true,
  "id": "rule_abc123"
}

POST/api/security/rules/:id/test

Test a rule against sample input to validate pattern matching before deployment.

Request Body

bash

curl -X POST "https://api.neuronedge.ai/api/security/rules/rule_abc123/test" \
  -H "Authorization: Bearer <api-key>" \
  -H "Content-Type: application/json" \
  -d '{
    "input": "SELECT * FROM users WHERE id = 1 UNION SELECT password FROM admin"
  }'

# Response
{
  "matched": true,
  "category": "prompt_injection",
  "action": "block",
  "confidence": 0.95,
  "matched_pattern": "(?i)(union|select|insert|drop|delete)\\s+(from|into|table)"
}

Guardrail Presets

GET/api/security/presets

List available rule presets (OWASP LLM Top 10, PII compliance, etc.).

Request Body

bash

curl -X GET "https://api.neuronedge.ai/api/security/presets" \
  -H "Authorization: Bearer <api-key>"

# Response
{
  "presets": [
    {
      "id": "preset_owasp_llm",
      "name": "OWASP LLM Top 10",
      "description": "Comprehensive protection against OWASP LLM Top 10 vulnerabilities",
      "categories": [
        "prompt_injection",
        "data_leakage",
        "model_denial",
        "supply_chain"
      ],
      "rule_count": 87
    },
    {
      "id": "preset_pii_compliance",
      "name": "PII Compliance",
      "description": "GDPR and CCPA compliant PII detection and redaction",
      "categories": ["pii_leakage"],
      "rule_count": 42
    }
  ]
}

POST/api/security/from-preset

Apply a preset to a policy, creating all associated rules.

Request Body

bash

curl -X POST "https://api.neuronedge.ai/api/security/from-preset" \
  -H "Authorization: Bearer <api-key>" \
  -H "Content-Type: application/json" \
  -d '{
    "preset_id": "preset_owasp_llm",
    "policy_id": "pol_production"
  }'

# Response
{
  "rules_created": 87,
  "policy_updated": true,
  "preset_id": "preset_owasp_llm",
  "policy_id": "pol_production"
}

Security Events

GET/api/security/events

List security events with filtering by type, category, severity, action, and time range.

Request

bash

curl -X GET "https://api.neuronedge.ai/api/security/events?severity=high&limit=10" \
  -H "Authorization: Bearer <api-key>"

# Response
{
  "events": [
    {
      "id": "evt_abc123",
      "type": "guardrail_triggered",
      "category": "prompt_injection",
      "severity": "high",
      "action": "block",
      "rule_id": "rule_abc123",
      "request_id": "01JKX8G7M2PQRSTUVWXYZ12345",
      "timestamp": "2026-03-10T14:25:30Z",
      "metadata": {
        "matched_pattern": "SQL injection attempt",
        "customer_id": "cust_xyz789"
      }
    }
  ],
  "total": 1247,
  "limit": 10,
  "offset": 0
}

GET/api/security/events/stats

Aggregate statistics for security events over a specified period.

Request

bash

curl -X GET "https://api.neuronedge.ai/api/security/events/stats?from=2026-03-01&to=2026-03-10" \
  -H "Authorization: Bearer <api-key>"

# Response
{
  "period": {
    "from": "2026-03-01T00:00:00Z",
    "to": "2026-03-10T23:59:59Z"
  },
  "total_events": 5432,
  "by_category": {
    "prompt_injection": 2100,
    "pii_leakage": 1800,
    "data_leakage": 892,
    "jailbreak": 640
  },
  "by_action": {
    "block": 3200,
    "redact": 1800,
    "flag": 432
  },
  "by_severity": {
    "critical": 120,
    "high": 890,
    "medium": 2400,
    "low": 2022
  }
}

GET/api/security/events/timeline

Event counts over time with configurable interval (hour or day).

Request

bash

curl -X GET "https://api.neuronedge.ai/api/security/events/timeline?interval=day&from=2026-03-01&to=2026-03-10" \
  -H "Authorization: Bearer <api-key>"

# Response
{
  "interval": "day",
  "data": [
    {
      "timestamp": "2026-03-01T00:00:00Z",
      "count": 542
    },
    {
      "timestamp": "2026-03-02T00:00:00Z",
      "count": 678
    },
    {
      "timestamp": "2026-03-03T00:00:00Z",
      "count": 590
    }
  ]
}

GET/api/security/events/top-patterns

Most frequently triggered attack patterns across all events.

Request

bash

curl -X GET "https://api.neuronedge.ai/api/security/events/top-patterns?limit=5" \
  -H "Authorization: Bearer <api-key>"

# Response
{
  "patterns": [
    {
      "pattern": "SQL injection via UNION",
      "category": "prompt_injection",
      "count": 1247,
      "percentage": 22.96
    },
    {
      "pattern": "SSN leakage",
      "category": "pii_leakage",
      "count": 890,
      "percentage": 16.38
    },
    {
      "pattern": "Jailbreak via role-play",
      "category": "jailbreak",
      "count": 640,
      "percentage": 11.78
    }
  ],
  "total_events": 5432
}

GET/api/security/events/posture

Overall security posture score (0-100) with contributing factors and recommendations.

Request

bash

curl -X GET "https://api.neuronedge.ai/api/security/events/posture?from=2026-03-01&to=2026-03-10" \
  -H "Authorization: Bearer <api-key>"

# Response
{
  "score": 87,
  "grade": "B+",
  "factors": {
    "coverage": 92,
    "response_effectiveness": 88,
    "incident_rate": 82,
    "policy_completeness": 85
  },
  "recommendations": [
    "Enable response-side PII detection for critical workflows",
    "Review and update jailbreak rules based on recent patterns"
  ],
  "period": {
    "from": "2026-03-01T00:00:00Z",
    "to": "2026-03-10T23:59:59Z"
  }
}

Red Team Scanning

POST/api/security/red-team/scan

Start a red team scan with specified intensity and target policy.

Request

bash

curl -X POST "https://api.neuronedge.ai/api/security/red-team/scan" \
  -H "Authorization: Bearer <api-key>" \
  -H "Content-Type: application/json" \
  -d '{
    "intensity": "comprehensive",
    "target_policy_id": "pol_production",
    "categories": ["prompt_injection", "jailbreak", "pii_leakage"]
  }'

# Response
{
  "scan_id": "scan_abc123",
  "status": "running",
  "intensity": "comprehensive",
  "target_policy_id": "pol_production",
  "categories": ["prompt_injection", "jailbreak", "pii_leakage"],
  "estimated_duration_minutes": 15,
  "started_at": "2026-03-10T14:40:00Z"
}

GET/api/security/red-team/reports

List all red team scan reports with status and summary metrics.

Request

bash

curl -X GET "https://api.neuronedge.ai/api/security/red-team/reports?limit=10" \
  -H "Authorization: Bearer <api-key>"

# Response
{
  "reports": [
    {
      "id": "scan_abc123",
      "status": "completed",
      "intensity": "comprehensive",
      "target_policy_id": "pol_production",
      "score": 87,
      "grade": "B+",
      "probes_run": 487,
      "vulnerabilities_found": 23,
      "started_at": "2026-03-10T14:40:00Z",
      "completed_at": "2026-03-10T14:55:30Z"
    }
  ],
  "total": 42,
  "limit": 10,
  "offset": 0
}

GET/api/security/red-team/reports/:id

Get full report including category breakdown, vulnerabilities, and recommendations.

Request

bash

curl -X GET "https://api.neuronedge.ai/api/security/red-team/reports/scan_abc123" \
  -H "Authorization: Bearer <api-key>"

# Response
{
  "id": "scan_abc123",
  "status": "completed",
  "score": 87,
  "grade": "B+",
  "intensity": "comprehensive",
  "target_policy_id": "pol_production",
  "probes_run": 487,
  "vulnerabilities_found": 23,
  "by_category": {
    "prompt_injection": {
      "probes": 150,
      "passed": 138,
      "failed": 12,
      "pass_rate": 92.0
    },
    "jailbreak": {
      "probes": 200,
      "passed": 189,
      "failed": 11,
      "pass_rate": 94.5
    },
    "pii_leakage": {
      "probes": 137,
      "passed": 137,
      "failed": 0,
      "pass_rate": 100.0
    }
  },
  "weaknesses": [
    {
      "category": "prompt_injection",
      "severity": "medium",
      "description": "Vulnerable to multi-step indirect prompt injection",
      "affected_probes": 8,
      "recommendation": "Enable context-aware prompt analysis"
    }
  ],
  "started_at": "2026-03-10T14:40:00Z",
  "completed_at": "2026-03-10T14:55:30Z"
}

GET/api/security/red-team/reports/:id/probes

Get individual probe results with filtering by category and pass/fail status.

Request

bash

curl -X GET "https://api.neuronedge.ai/api/security/red-team/reports/scan_abc123/probes?category=prompt_injection&passed=false" \
  -H "Authorization: Bearer <api-key>"

# Response
{
  "probes": [
    {
      "id": "probe_xyz789",
      "category": "prompt_injection",
      "technique": "Multi-step indirect injection",
      "input": "Ignore previous instructions...",
      "expected_action": "block",
      "actual_action": "allow",
      "passed": false,
      "severity": "medium",
      "timestamp": "2026-03-10T14:42:15Z"
    }
  ],
  "total": 12,
  "limit": 20,
  "offset": 0
}

Error Responses

Status	Description
400	Invalid request parameters
401	Missing or invalid API key
403	Insufficient tier / feature not available
404	Resource not found
429	Rate limit exceeded

400 Bad Request

Response

json

{
  "error": {
    "code": "invalid_request",
    "message": "Invalid request parameters",
    "details": {
      "category": "Unknown category: invalid_category. Valid: prompt_injection, jailbreak, pii_leakage, data_leakage, model_denial"
    }
  }
}

401 Unauthorized

Response

json

{
  "error": {
    "code": "unauthorized",
    "message": "Missing or invalid API key",
    "details": "Authorization header must include a valid Bearer token"
  }
}

403 Forbidden

Response

json

{
  "error": {
    "code": "forbidden",
    "message": "Insufficient tier / feature not available",
    "details": "Security API requires Professional tier or above. Current tier: Starter"
  }
}

404 Not Found

Response

json

{
  "error": {
    "code": "not_found",
    "message": "Resource not found",
    "details": "Rule with id 'rule_invalid' does not exist"
  }
}

429 Rate Limit Exceeded

Response

json

{
  "error": {
    "code": "rate_limit_exceeded",
    "message": "Rate limit exceeded",
    "details": "Security API limit: 100 requests/minute. Retry after 42 seconds.",
    "retry_after": 42
  }
}

Rate Limits

Security API endpoints have tier-specific rate limits:

Tier	Security API Limit
Professional	100 requests/minute
Enterprise	500 requests/minute

Rate limits are enforced per API key. When exceeded, the API returns a 429 status with a retry_after value in seconds.