🧠 SynthGuard™ User Guide

# SynthGuard (shadow-ai) — User Guide

**Port:** 8021 | **Category:** ai | **Caps:** 30

AI/ML Security — model protection, adversarial defense, AI governance, prompt injection detection

## Keyboard Shortcuts
/ search | t triage | m monitor | i investigate | a assistant | ? help | Cmd+K palette

## Capabilities (30)
- model_protection
- adversarial_defense
- ai_governance
- prompt_injection_detection
- model_drift_detection
- training_data_protection
- llm_security
- supply_chain_ml
- data_poisoning_detection
- model_watermarking
- inference_attack_detection
- federated_learning_security
- ... and 18 more

## Actions (12)
- block_prompt
- quarantine_model
- create_alert
- rollback_model
- scan_model
- audit_pipeline
- generate_abom
- disable_model_endpoint

## AI Features
- PounceSOC™ (auto-triage)
- PounceForge™ (agentic investigation)
- WraithMind™ (AI assistant)
- ShadowNeural™ (ML detection)

## Managed Services
- managed AI security
- AI red teaming
- model audit
- ai governance operations
- llm red teaming retainer
- model release assurance