CyberAg Documentation

CyberAg Documentation

Enterprise-grade security for AI agents: discovery, red teaming, guardrails, NHI governance, Replay Engine, Kill Switch, and Audit Vault.

Getting started

GuideDescription
Local SetupRun the full stack on your machine with Docker + LocalStack
ArchitectureSystem design, services, data model, and event flow
API ReferenceREST endpoints for console, SDK, and integrations
Test DataSeeded sample data and how to reset it
TestingUnit, Storybook, and E2E test strategy

Planning & compliance

DocumentDescription
Project PlanPhases, milestones, module specs, acceptance criteria
Threat ModelSecurity controls and threat mapping
Load Test PlanPerformance targets and k6 scenarios
Component PlanDesign system and console adoption status

Runbooks

RunbookDescription
Local DevelopmentDay-to-day local workflow, verify, reset
Kill SwitchActivate and verify fleet-wide stop
Connector OnboardingAdd and sync estate connectors

Integrations

IntegrationLocation
TypeScript SDKpackages/sdk — see API Reference
Python SDKintegrations/python — LangChain, CrewAI, LangGraph, AutoGen
WrapSec gatewayintegrations/README.md
OpenClaw pluginintegrations/openclaw

Applications

AppURL (local)URL (production)
Consolehttp://localhost:3002https://dashboard.cyberag.io
Docs sitehttp://localhost:3003https://docs.cyberag.io
Marketinghttp://localhost:3000https://cyberag.io

Quick commands


npm install
npm run start:local          # LocalStack + deploy + seed + console
npm run local:verify         # Smoke test API + Gemini features
npm run test                 # Unit tests
npm run test:e2e             # Playwright E2E (requires local stack)
npm run storybook            # Design system at :6006

Read-only local UI (no Gemini key required):


$env:CYBERAG_LOCAL_SKIP_GEMINI="1"
npm run start:local