How openbas-platform/openbas works
Project type & purpose
• This is a full-stack, modular platform (“OpenBAS”) for planning and running cyber-adversary simulations.
• It comprises:
• A Java Spring Boot backend (“openbas-api”)
• A React/TypeScript front-end SPA (“openbas-front”)
• A shared Java model layer (“openbas-model”)
• A framework module for common code (“openbas-framework”)
• Dev orchestration configs (Docker, docker-compose, CI/CD)
File-structure insights
• Top-level modules:
• openbas-api: monolithic REST API + scheduler + collectors + plugin-style executors/injectors
• openbas-front: React UI (public & private/admin), E2E tests
• openbas-model: data model + persistence + ES engine + drivers
• openbas-framework: shared abstractions (queues, contracts)
• openbas-dev: docker-compose definitions for multi-container local/development topology
• Backend layering:
• Controller (rest package) → Service → Repository (JPA/SQL, ES)
• AOP, security, caching, scheduling modules
• Asynchronous work via RabbitMQ (queue configs & helpers)
• Executors/injectors connect to external platforms (Caldera, CrowdStrike, email, SMS, Mastodon, OpenCTI…)
• Infra/config:
• application.properties, Flyway migrations, OTLP telemetry, Pyroscope
• Dockerfiles for containers, GitHub Actions/Drone/CircleCI
Key README points
• Platform edition (Community vs. Enterprise)
• Architecture: core modules + integrations with OpenCTI, message platforms, OpenStreetMap
• Deployment via Docker or manual (db, ES, RabbitMQ, Minio, OTLP collector)
• Front-end (React SPA + NGINX or static-file server)
• API Backend (Spring Boot monolith)
• Relational DB (PostgreSQL/MySQL)
• Search Engine (Elasticsearch)
• Object Store (Minio/S3)
• Message Broker (RabbitMQ)
• Telemetry Collector (OTel Collector + Pyroscope)
• External services:
– OpenCTI (threat intel)
– SMTP server (email)
– SMS provider (OVH)
– Social platform (Mastodon)
• Executors/Injectors (plugin workers)
• Browser → Front-end (HTTP/S)
• Front-end → API (REST calls, WebSocket/SSE for live updates)
• API ↔ DB (JPA)
• API ↔ ES via openbas-model (for full-text search)
• API ↔ Minio (file uploads: agents, implants)
• API → RabbitMQ (publish tasks & scheduling triggers)
• Executors subscribe to RabbitMQ → perform platform-specific work → send results back via API or RabbitMQ
• API → Telemetry Collector (OTLP) → Pyroscope storage
• API ↔ External services (OpenCTI, SMTP, SMS, Mastodon)
c. Architectural patterns & principles:
• Layered monolith (Controller→Service→Repository)
• Modular design with plugin-style executors/injectors
• Event-driven asynchronous tasks over RabbitMQ
• Shared model and framework modules reused across API and executors
• Configuration-driven (Spring profiles, Docker)
d. Technologies & frameworks:
• Front-end: React, Redux, Yarn v4, Vite/Vitest, Playwright
• Backend: Java 17+, Spring Boot, Spring Security, Flyway, AOP, OpenTelemetry
• DB: SQL (Postgres/MySQL), JPA/Hibernate
• Search: Elasticsearch via custom driver
• Object storage: Minio (S3 API)
• Broker: RabbitMQ
• Telemetry: OTel Collector, Pyroscope
• CI/CD: GitHub Actions, Drone, CircleCI, Docker