🐉
Select an incident or trigger a new War Room
0/8 up
5:05:06 PMPlatform Monitoring↓ Dragon governance · Playground · Auth risk engine · Signal integrity · Compliance5:05:06 PM
🎯
Dragon Command Center
Live Dragon governance feed — fire demo alerts, monitor active incidents, review the audit trail.
🎮 Fire a Live Alert
Creates a live P1/P2 alert in Dragon. Appears in the feed below within seconds.
Active Alerts
No active alerts
Use the fire buttons above
Governance Decisions
Audit TrailNo governance decisions yet
🐉
Dragon Playground
6 real governance scenarios. Each fires a live risk assessment through Dragon's policy engine.
🧪Run Test Suite
AUTO-APPROVEDpytest tests/integration/Test environment only
✨Code Formatting
AUTO-APPROVEDnpm run formatStyle changes only
⚠️Production Deploy
HUMAN APPROVALterraform apply -target=production -auto-approveFull prod infra update
💰Pricing Change
BLOCKEDkubectl set env deployment/api PRICE_MULTIPLIER=1.30$500K revenue impact
🔐IAM Escalation
HARD BLOCKaws iam attach-user-policy --user-name intern-bob --policy-arn ...AdministratorAccessFull admin access to AWS
💣Database Delete
HARD BLOCKDROP DATABASE production_customers;Permanent data deletion
🔐
KlynxAuth Risk Engine
15-signal weighted risk classifier on every authentication event. TOR detection, impossible travel, device fingerprinting.
tor_detected
impossible_travel
vpn_detected
unknown_device
atypical_country
atypical_hour
high_velocity
new_asn
days_since_login
ip_reputation
✅ALLOW
Normal Login
Typical weekday login from known device, home country, business hours
Risk Score12/100
⚠️BLOCK
TOR Exit Node
Login from TOR exit node, new unrecognized device, atypical country
Risk Score88/100
🌍BLOCK
Impossible Travel
User signed in from NYC 28 min ago — now attempting from Moscow
Risk Score96/100
🔑CHALLENGE
Step-Up Challenge
Known user from known device but logging in at 3am — triggers MFA
Risk Score42/100
🔐
TOTP / FIDO2
RFC 6238 + WebAuthn
📋
SCIM 2.0
Enterprise provisioning
🔑
OAuth 2.0 / OIDC
SSO + token flows
🧠
ML Risk Engine
15-signal classifier
🏗
Platform Health
Live status across all Klynx AI services. Auto-checks every 60 seconds.
Dragon Core
AI Governance Engine
Checking…
DevOps Orchestrator
RCA + Autofix + Playground
Checking…
KlynxAuth
Identity + MFA Platform
Checking…
KlynxBot
19-Domain AI Assistant
Checking…
KlynxChip
Edge AI Inference Runtime
Checking…
KlynxGuard
API Security Layer
Checking…
KlynxCortex
AGI Memory + Reflection
Checking…
Marketing Site
klynxai.com
Checking…
📡
Signal Integrity Monitor
Data Confidence & Context Integrity family — the Layer 1→2 gate. Determines whether downstream inference is allowed to run.
Loading signal integrity data…
🔏 SDL Audit Explorer
KlynxChip Secure Decision Ledger — Ed25519 + SHA-256 chain integrity
Chain Status
—Entries Verified
—
Ed25519 Signatures
—
ECDSA P-256 · offline verifiable
Click Verify Chain to audit the KlynxChip decision ledger
📋 Policy Control Center
YAML Policy Engine — hot-reload · <1ms evaluation · 6 action types
Live Policy Evaluation
📤 Upload YAML Policy
✅ Compliance Coverage Report
Automated coverage derived from deployed PGE policies · 15 active policies
NIST AI RMF
Full lifecycle AI risk governance
GOVERN 1.1·GOVERN 1.7
15 policiesSOX Compliance
Financial AI dual-approval controls
Section 302·Section 404
2 policiesHIPAA
PHI shadow-routing to local model
Privacy Rule·Security Rule
1 policyGDPR / EU Privacy
Cross-border data transfer blocks
Art. 28 Processor·Art. 46 Cross-border
2 policiesEU AI Act
High-risk AI system controls
Art. 9 Risk Mgmt·Art. 13 Transparency
4 policiesCMMC 2.0
DoD supply chain model controls
CM.3.068·AC.2.006
2 policiesOverall Compliance Score
Weighted average across 6 regulatory frameworks
78%
26 controls mapped
🇪🇺 EU AI Act — Article-Level Mapping
Regulation EU 2024/1689 · 12 articles assessed
🚦 Human Escalation Queue
Dragon decisions requiring human review before execution
✅
No pending escalations
Dragon policies have not flagged any decisions requiring human review
Escalations appear here when Dragon PGE fires an ESCALATE_TO_HUMAN policy action
👁 AI Visibility — Asset Inventory
Every AI model, agent, API, and connector in your estate · Shadow AI detection
—
Total Assets
—
Ungoverned
—
Gov. Coverage
0
Shadow AI
✅ No shadow AI detected
Register AI assets via POST /api/visibility/assets/register
🤖 MCP Security Gateway
AI agent tool call governance · Kill switch · Full audit log
🤖
No agent sessions yet
Sessions appear when AI agents call POST /api/mcp/intercept
🛡 AI Security Posture (AI-SPM)
Scan AI endpoints for misconfigurations · TLS · Auth · CORS · Rate limiting · Dragon coverage
No scans yet. Enter an AI endpoint URL above and click Scan.
🎯 AI Red Team — Security Testing
54 attack payloads · Prompt injection · Jailbreak · PII exfil · System leak · Credentials
🎯
No scans yet
Enter an AI endpoint URL above — we send 54 attack payloads and score vulnerability
🚨 AI Detection & Response (AIDR)
Auto-detects: repeated gate failures · high MCP block rate · KlynxChip crisis mode
🛡
No incidents detected
Click Run Detection to scan for anomalies across Dragon, MCP Gateway, and KlynxChip
👤 Shadow AI Discovery
Detect unauthorized AI tool usage across your org — ungoverned ChatGPT, Claude, Copilot, Gemini
👤
No shadow AI data yet
Connect your proxy or network gateway to start reporting AI tool usage via POST /api/shadow/report
💰 AI Budget Enforcement
Per-org, per-model token/cost limits. Dragon blocks requests when limits are exceeded.
💰
No budget data yet
Set budgets via POST /api/budget/set and record usage via POST /api/budget/record
🛡 Prompt Injection Shield
Real-time detection of role overrides, jailbreaks, indirect injection, and data exfiltration attempts
🛡
No scan data yet
Send prompts via POST /api/shield/scan
👻 KlynxGhost — Data Masking
Auto-redact PII, PHI, API keys, credentials from AI inputs — GDPR / HIPAA compliance layer
👻
No masking data yet
Call POST /api/ghost/mask before sending user input to any LLM
🔑 KlynxVault — Secrets Manager
Encrypted storage for API keys, model credentials, system prompts. Every access logged.
🔑
No secrets stored yet
Store API keys and credentials via POST /api/vault/secrets
📋 One-Click Compliance Export
Auto-generate SOC2 / EU AI Act / ISO 42001 / HIPAA / CMMC audit reports from live Klynx data
⚡ AI Circuit Breaker
Real-time model health — auto-opens when failure rate exceeds threshold
No circuit data yet — circuits are created on first LLM call
🧠 Multi-Model Consensus
Parallel LLM evaluation — Dragon picks the safest, highest-quality response
No consensus evaluations yet — call POST /api/consensus/evaluate
📰 KlynxCast — AI Governance Digest
Claude-generated executive summary of platform activity, risks, and wins
Click "Generate Now" to create an AI-powered executive digest
🧪 Policy Simulator
Test Dragon policies in sandbox — validate before deploying to production
💚 Platform Health Score
Weighted 0-100 score across security, reliability, compliance, and governance
Health score is computed from live module data — loading…
📊 Behavioral Baseline + Anomaly Detection
Statistical profiling of AI usage — flags prompt spikes, unusual hours, burst traffic, model switches
No baseline data yet — events are recorded via POST /api/baseline/event
🛒 Policy Marketplace
Browse All ↗Loading marketplace policies…
0/8 services operational · Auto-refreshes every 30s
📣 Social Publisher
Generate a post from a brief — Claude drafts, you approve before publishing.
No pending drafts. Write a brief above to generate your first post.