How kashyapharshit/cybersi works
CYBERSI is a security intelligence and risk-management platform. Security evidence is ingested and normalized by the Node backend, persisted in MongoDB, analyzed for asset and portfolio risk by the FastAPI intelligence service, and presented in role-based workspaces. The platform also supports analyst queries, control-investment optimization, regulatory coverage, and audit integrity. The graph follows documented system boundaries and sampled call paths; several domain subsystems are represented at their existing directories because their implementations were not sampled.
Main entry points: app.js HTTP API · main.py FastAPI engine · ingestion.service.js Evidence ingestion · models Evidence store
Components 25
User experience
- Role workspaces
attack_on_point_frontend_redesigned/src/App.jsx - API client
attack_on_point_frontend_redesigned/src/api.js
Evidence backend
- HTTP API
attack_on_point_backend_hardened/src/app.js - Evidence ingestion
attack_on_point_backend_hardened/src/services/ingestion.service.js - Source adapters
attack_on_point_backend_hardened/src/parsers - Domain services
attack_on_point_backend_hardened/src/services - Ingestion jobs
attack_on_point_backend_hardened/src/models/IngestionJob.js - Evidence store
attack_on_point_backend_hardened/src/models - Regulatory coverage
attack_on_point_backend_hardened/src/services/regulatoryCoverage.service.js
Risk intelligence
- Analysis endpoints
attack_on_point_backend_hardened/src/routes/ai.routes.js - AI result handling
attack_on_point_backend_hardened/src/controllers/ai.controller.js - Analysis trigger
attack_on_point_backend_hardened/src/services/aiTrigger.service.js - FastAPI engine
attack_on_point_ai_fastapi_hardened/app/main.py - Backend connector
attack_on_point_ai_fastapi_hardened/app/backend_client.py - Risk calculations
attack_on_point_ai_fastapi_hardened/app/math_engine.py - Control optimizer
attack_on_point_ai_fastapi_hardened/app/optimizer.py - Analyst copilot
attack_on_point_ai_fastapi_hardened/app/llm.py - Query forwarding
attack_on_point_backend_hardened/src/services/analystQuery.service.js
Audit integrity
- Audit and anchoring
attack_on_point_backend_hardened/src/services/blockchain.service.js - Risk audit records
attack_on_point_backend_hardened/src/models/RiskAudit.js
Other components
- Platform users
- MongoDB
- Security sources
- Ollama
- Blockchain network
Connections 28
- Platform users → Role workspaces: uses
- Role workspaces → API client: requests data
- API client → HTTP API: calls API
- Security sources → Evidence ingestion: supplies evidence
- Evidence ingestion → Source adapters: parses findings
- Evidence ingestion → Ingestion jobs: records job
- Evidence ingestion → Evidence store: stores evidence
- HTTP API → Analysis endpoints: dispatches requests
- Analysis endpoints → Analysis trigger: triggers analysis
- Analysis endpoints → AI result handling: handles results
- Analysis trigger → FastAPI engine: requests analysis
- FastAPI engine → Backend connector: fetches and submits
- Backend connector → HTTP API: exchanges payloads
- FastAPI engine → Risk calculations: calculates risk
- FastAPI engine → Control optimizer: optimizes controls
- FastAPI engine → Analyst copilot: generates summaries
- Analyst copilot → Ollama: queries model
- HTTP API → Query forwarding: dispatches query
- Query forwarding → FastAPI engine: forwards query
- HTTP API → Domain services: dispatches domain work
- Domain services → Evidence store: reads and writes
- Regulatory coverage → Evidence store: reads evidence
- HTTP API → Audit and anchoring: verifies reports
- Audit and anchoring → Risk audit records: reads audit records
- Audit and anchoring → Blockchain network: anchors hashes
- Risk audit records → MongoDB: persists records
- Evidence store → MongoDB: persists evidence
- Ingestion jobs → MongoDB: persists jobs