AlertForge is a production-grade incident management and response orchestration platform. Designed for modern engineering teams, it bridges the gap between infrastructure monitoring and actionable resolution. With AI-driven postmortems, real-time collaboration "War Rooms," and multi-channel notifications, AlertForge ensures your team spends less time panic-searching and more time resolving.
Stop writing post-incident reports manually. AlertForge uses a state-of-the-art LangGraph pipeline to ingest incident timelines, chat logs, and external technical insights to generate high-fidelity, hallucination-free postmortems automatically upon resolution.
Collaborate in live, purpose-built environments. Features include:
- Live Chat & Task Checklists: Synchronized across all responders via Socket.IO.
- Presence Tracking: See exactly who is investigating in real-time.
- Auto-Timeline: Every action is automatically logged to the immutable audit trail.
Never miss a critical pulse. Route alerts dynamically to:
- Telegram & Discord Webhooks.
- WhatsApp (via Twilio).
- Email (SMTP/Nodemailer).
- Custom Webhooks for external integrations.
Strict organization-level data isolation. Manage multiple teams, services, and environments under one unified dashboard with robust Role-Based Access Control (RBAC).
A unified catalog of your infrastructure. Monitor uptime percentages and see the cascading impact of incidents on downstream services in real-time.
- Core: React 19 + Vite (Ultra-fast HMR)
- Styling: Tailwind CSS 4 + Shadcn UI
- Animations: Framer Motion + GSAP (For a premium, interactive UI)
- Auth: Clerk (Secure User Management)
- State/Data: Axios + Socket.io-client
- Server: Node.js + Express 5
- Database: MongoDB (Mongoose) + Redis (ioredis)
- Real-time: Socket.IO with Redis Adapter (Horizontally scalable)
- AI/ML: LangChain + LangGraph + Mistral AI
- Queues: BullMQ for resilient background processing
AlertForge follows a strictly Layered Architecture to ensure maintainability and production resilience.
flowchart TD
Client["User/SDK Client"] -- HTTP/Socket --> Auth["smartAuth Middleware"]
Auth -- Verified --> RBAC["RBAC Middleware"]
RBAC -- Authorized --> Controller["Controller Layer"]
Controller -- Validate & Call --> Service["Service Layer (Orchestration)"]
subgraph "The Orchestration Layer"
Service -- Transactional --> DAO["DAO Layer"]
DAO -- Query --> DB[("MongoDB")]
Service -- Side Effect --> Timeline["Timeline Service"]
Service -- Side Effect --> Socket["Socket Service (Real-time)"]
Service -- Side Effect --> Notify["Notification Service"]
Service -- Side Effect --> AI["LangGraph Postmortem"]
end
Service -- Response --> Controller
Controller -- ApiResponse --> Client
- Node.js (v18+)
- MongoDB Instance
- Redis Instance (Upstash recommended)
- API Keys: Mistral AI, Twilio, Clerk, Tavily
-
Clone the Repository
git clone https://github.com/Soumipal56/AlertForge.git cd AlertForge/AlertForge/AlertForge -
Setup Backend
cd Backend npm install cp .env.example .env # Update your variables npm run dev
-
Setup Frontend
cd ../Frontend npm install cp .env.example .env # Update your variables npm run dev
- Dual-Mode Auth: Seamlessly switch between Dashboard (Cookie) and SDK (API Key) auth.
- Distributed Blacklisting: Redis-backed JWT invalidation for instant session termination.
- Database Indexing: Optimized compound indexes for O(1) dashboard aggregations.
- Atomic Operations: MongoDB transactions ensure data integrity across incidents and timelines.
Distributed under the ISC License. See LICENSE for more information.
"Forging the future of SRE tools."