How shreekamesh/black_veil works
BLACK VEIL is an autonomous cognitive cyber-defense framework. Its documented prediction workflow accepts a model and feature vector through the inference API and returns a prediction; the broader defense workflow evaluates threats and trust, reasons about response, and can deploy deception or recovery. The repository contains overlapping API and engine generations, so the graph emphasizes README-described capabilities and source-index-supported links. Several subsystem connections are documented rather than visible in the sampled implementation excerpts.
Main entry points: acdo.py Defense Orchestrator · model_loader.py Model Loader · inference_api.py Inference API · cognitive Cognitive Reasoning
Components 21
Product Interfaces
- Inference API
inference_api.py - Web Dashboard
frontend/src/App.tsx - API Application
api/app.py - Security Reports
report_engine/engine.py
Threat Intelligence
- Domain Detection Models
ai_core - Threat Knowledge
knowledge_engine/engine.py - Forensic Analysis
forensic_engine/engine.py - Attack Memory Graph
attack_memory_graph/engine.py
Cognitive Defense
- Defense Orchestrator
src/core/orchestrator/acdo.py - Cognitive Reasoning
src/core/cognitive - Decision Consensus
ai_decision_brain/engine.py - Ethics Layer
ai_ethics_layer/engine.py - Trust Evaluation
trust_engine/engine.py
Adaptive Response
- Trust Recovery
temporal_recovery_engine/engine.py - Deception Engine
deception_engine/engine.py - Deception Evolution
deception_evolution_engine/engine.py - Security Actions
src/core/security - Self-Healing
self_healing_engine/engine.py
Model and State
- Model Loader
ai_core/model_loader.py - Model Registry
src/ai_foundation/core/model_registry.py
Other components
- Operator
Connections 21
- Operator → Inference API: submits features
README.md - Inference API → Model Loader: loads models
- Inference API → Operator: returns prediction
README.md - Operator → Web Dashboard: uses dashboard
- Operator → API Application: accesses API
- API Application → Model Loader: imports loader
api/app.py - Model Loader → Domain Detection Models: serves models
ai_core/cicids_engine/engine.py - Defense Orchestrator → Cognitive Reasoning: coordinates defense
README.md - Defense Orchestrator → Trust Evaluation: coordinates trust
README.md - Defense Orchestrator → Deception Engine: coordinates deception
README.md - Defense Orchestrator → Attack Memory Graph: coordinates memory
README.md - Defense Orchestrator → Security Actions: coordinates actions
README.md - Cognitive Reasoning → Decision Consensus: supports decisions
- Cognitive Reasoning → Ethics Layer: checks decisions
- Trust Evaluation → Trust Recovery: uses recovery
trust_engine/engine.py - Deception Evolution → Deception Engine: evolves strategy
- Forensic Analysis → Attack Memory Graph: records findings
- Threat Knowledge → Defense Orchestrator: informs defense
- Defense Orchestrator → Self-Healing: coordinates recovery
- Defense Orchestrator → Security Reports: requests summaries
- Model Registry → Model Loader: registers models