Open-source AI Security Operations Center: alert fusion, LLM-agent triage, MITRE ATT&CK investigation, and a replayable decision ledger for every agent step. Self-hostable, runs with no API keys, MIT licensed. Ships an MCP server for Claude, Cursor and Continue.
Python agentic-aiai-securityalert-triageblue-teamcybersecuritydetection-engineeringincident-responselanggraphllm-agentsmcpmitre-attackmodel-context-protocolpurple-teamsecurity-operationsself-hostedsiemsoarsocthreat-detectionthreat-intelligence
/**
* Multi-tenancy and RBAC types
*/
export type UserRole =
| "platform_admin" // Manages the platform itself
| "tenant_admin" // Full access within a tenant
| "security_manager" // Read all, approve actions
| "analyst_tier1" // Triage, acknowledge alerts
| "analyst_tier2" // Full investigation, create cases
| "analyst_tier3" // Advanced hunting, custom rules
| "threat_hunter" // Dedicated threat hunting
| "auditor" // Read-only, compliance reporting
| "readonly"; // Read-only general
export type TenantPlan = "starter" | "professional" | "enterprise" | "mssp";
export interface TenantLimits {
events_per_day: number;
connectors: number;
retention_days: number;
users: number;
ai_queries_per_day: number;
playbooks: number;
custom_rules: number;
}
export interface TenantSettings {
mfa_required: boolean;
sso_enabled: boolean;
sso_provider?: string;
ip_allowlist?: string[];
data_residency_region?: string; // "us", "eu", "ap"
encryption_key_id?: string; // Customer-managed encryption key
auto_assign_alerts?: boolean;
default_alert_sla_hours?: number;
slack_webhook?: string;
pagerduty_integration_key?: string;
notifications?: {
email?: string[];
slack_channel?: string;
critical_only?: boolean;
};
}
export interface Tenant {
id: string;
name: string;
slug: string;
plan: TenantPlan;
status: "active" | "suspended" | "trial" | "churned";
created_at: string;
trial_ends_at?: string;
limits: TenantLimits;
settings: TenantSettings;
industry?: string;
country?: string;
contact_email?: string;
}
export interface User {
id: string;
tenant_id: string;
email: string;
name: string;
role: UserRole;
status: "active" | "invited" | "suspended";
created_at: string;
last_login?: string;
mfa_enabled: boolean;
preferences?: {
timezone?: string;
theme?: "dark" | "light" | "system";
notification_channels?: ("email" | "slack" | "pagerduty")[];
dashboard_layout?: string;
alert_auto_refresh_seconds?: number;
};
}
export interface ApiKey {
id: string;
tenant_id: string;
name: string;
key_prefix: string; // Only first 8 chars stored
scopes: string[];
created_by: string;
created_at: string;
expires_at?: string;
last_used?: string;
is_active: boolean;
}
/** RBAC permission check */
export interface PermissionCheck {
user_role: UserRole;
resource: string; // e.g., "alerts", "cases", "actions", "rules"
action: "read" | "write" | "delete" | "approve" | "admin";
}
export const ROLE_PERMISSIONS: Record<UserRole, Record<string, string[]>> = {
platform_admin: { "*": ["read", "write", "delete", "approve", "admin"] },
tenant_admin: { "*": ["read", "write", "delete", "approve"] },
security_manager: {
alerts: ["read", "write"],
cases: ["read", "write", "approve"],
actions: ["read", "approve"],
rules: ["read"],
reports: ["read", "write"],
users: ["read"],
},
analyst_tier1: {
alerts: ["read", "write"],
cases: ["read"],
actions: ["read"],
rules: ["read"],
reports: ["read"],
},
analyst_tier2: {
alerts: ["read", "write"],
cases: ["read", "write"],
actions: ["read", "write"],
rules: ["read"],
reports: ["read", "write"],
},
analyst_tier3: {
alerts: ["read", "write"],
cases: ["read", "write"],
actions: ["read", "write"],
rules: ["read", "write"],
reports: ["read", "write"],
playbooks: ["read", "write"],
},
threat_hunter: {
alerts: ["read", "write"],
cases: ["read", "write"],
actions: ["read"],
rules: ["read", "write"],
hunt: ["read", "write"],
reports: ["read", "write"],
},
auditor: { "*": ["read"] },
readonly: { "*": ["read"] },
};