Open Arsenal

See the stars

Umbrix Specification

Home / Umbrix / Umbrix Specification

Umbrix

Adaptive privacy for a connected world.

Umbrix is a commercial-grade, open-source privacy and network security platform designed to provide adaptive network obfuscation, multi-node routing, encrypted communications, application isolation, and per-agent network privacy across desktop, mobile, server, IoT, and locally hosted AI environments.

Umbrix is designed as a modular system. Core modules provide the foundational networking, cryptographic, routing, privacy, security, and multi-channel capabilities required by the platform. Optional plugin modules extend Umbrix with specialized functionality without requiring those capabilities to be installed or enabled.

A central design goal of Umbrix is to make network privacy available at the individual application and AI-agent level. Multiple AI agents can operate simultaneously while receiving independently managed Umbrix channels, routing policies, encryption sessions, DNS policies, privacy profiles, and network identities.

Project Goals

Umbrix is designed to:

  • Protect network traffic through authenticated encryption.
  • Mask public IP addresses from destination services.
  • Separate network origin from destination traffic.
  • Support configurable multi-node routing.
  • Provide adaptive network obfuscation.
  • Reduce exposure to traffic analysis.
  • Protect DNS and network metadata.
  • Provide application-specific network isolation.
  • Provide dedicated VPN channels for individual AI agents.
  • Support independently managed privacy policies for every channel.
  • Support locally hosted AI systems with network privacy by default.
  • Support self-hosted and distributed infrastructure.
  • Minimize unnecessary centralized dependencies.
  • Provide cryptographic agility.
  • Support desktop, mobile, server, IoT, and AI infrastructure.
  • Provide a modular plugin ecosystem.
  • Maintain transparent, auditable open-source architecture.

Umbrix treats privacy as a layered system rather than simply an encrypted tunnel. Network routing, transport behavior, DNS, applications, AI agents, device interfaces, metadata, and infrastructure are treated as separate privacy boundaries.

Design Principles

Privacy by Architecture

Privacy protections should be structural rather than dependent on users correctly configuring numerous individual settings.

Modular Design

Core capabilities should be separated into independently maintainable modules with explicit interfaces and security boundaries.

Core Security

Security-critical functionality belongs in the core platform and should not depend on optional plugins.

Optional Extensibility

Specialized capabilities should be implemented as optional plugins when they are not required for the fundamental operation of Umbrix.

Local First

Umbrix should support local and self-hosted operation wherever practical.

Distributed Infrastructure

Umbrix should support distributed nodes, decentralized discovery, and distributed routing without requiring a single centralized authority.

Agent Isolation

Every AI agent should be capable of receiving an independently managed network channel.

Application Isolation

Applications should be able to receive independent routing, DNS, encryption, and privacy policies.

Human Control

Automatic routing and security decisions should remain configurable and understandable.

Cryptographic Agility

Umbrix should support replacement and addition of cryptographic mechanisms as standards evolve.

Cross Platform

The architecture should maintain consistent privacy behavior across supported operating systems while respecting platform-specific security boundaries.

Measurable Privacy

Privacy claims should be evaluated through testing and measurable security properties rather than assumed from architectural intent.

Architecture

Umbrix consists of a modular privacy platform containing:

  • Client services
  • Tunnel management
  • Protocol services
  • Cryptographic services
  • Routing services
  • Node services
  • Control-plane services
  • Privacy policy services
  • Multi-channel services
  • Application isolation
  • AI-agent isolation
  • Security monitoring
  • Optional plugins
  • Developer APIs

Core modules provide foundational functionality.

Optional plugins extend the platform through defined interfaces and permission boundaries.

Core Modules

Core Client Module

The Client Module provides the primary local Umbrix service.

Capabilities include:

  • Connection management
  • Tunnel lifecycle management
  • Privacy profile management
  • Channel management
  • Route selection
  • Configuration management
  • Platform integration
  • Connection health monitoring
  • Automatic reconnection
  • Secure shutdown
  • Policy enforcement
  • Security state reporting

The client should not expose private cryptographic material through ordinary user interfaces.

Core Tunnel Module

The Tunnel Module provides the local encrypted network interface.

Capabilities include:

  • TUN/TAP integration
  • Platform-specific virtual interfaces
  • Packet encapsulation
  • Packet forwarding
  • Packet authentication
  • Packet reassembly
  • MTU management
  • Connection isolation
  • Traffic leak prevention
  • Tunnel lifecycle management

The Tunnel Module provides the foundation for device-wide and channel-specific routing.

Core Protocol Module

The Protocol Module provides the Umbrix protocol architecture.

Capabilities include:

  • Secure session establishment
  • Mutual authentication
  • Session negotiation
  • Multiplexed traffic
  • Ephemeral session keys
  • Forward secrecy
  • Routing envelopes
  • Packet authentication
  • Replay protection
  • Protocol version negotiation
  • Transport negotiation
  • Cryptographic negotiation
  • Session expiration
  • Session rotation

The protocol architecture should remain independent from individual network transports.

Core Cryptography Module

The Cryptography Module provides cryptographic primitives and key-management services.

Capabilities include:

  • X25519 key exchange
  • ChaCha20-Poly1305 authenticated encryption
  • BLAKE3 hashing
  • Cryptographically secure random generation
  • Ephemeral session keys
  • Per-channel keys
  • Per-hop keys
  • Forward secrecy
  • Key rotation
  • Replay protection
  • Secure key destruction
  • Cryptographic agility
  • Hybrid post-quantum key exchange support

Umbrix should use independently reviewed cryptographic libraries wherever practical.

Core Routing Module

The Routing Module manages traffic paths through Umbrix infrastructure.

Capabilities include:

  • Single-node routing
  • Multi-node routing
  • Configurable hop counts
  • Entry-node selection
  • Relay-node selection
  • Exit-node selection
  • Route health monitoring
  • Route expiration
  • Route rotation
  • Route replacement
  • Route failover
  • Privacy-aware routing
  • Performance-aware routing
  • Reliability-aware routing
  • Route diversity
  • Policy-based routing

Routing should minimize unnecessary knowledge of the complete path at individual nodes.

Core Adaptive Routing Module

The Adaptive Routing Module evaluates network conditions and determines whether routes remain appropriate.

It may consider:

  • Latency
  • Packet loss
  • Node availability
  • Congestion
  • Reliability
  • Route diversity
  • Transport health
  • Privacy requirements
  • Node reputation
  • Configured user policies

The module can replace routes when configured thresholds or privacy requirements are no longer satisfied.

Core Obfuscation Module

The Obfuscation Module provides foundational traffic transformation.

Capabilities include:

  • Packet padding
  • Packet-size normalization
  • Timing variation
  • Traffic batching
  • Flow shaping
  • Connection behavior normalization
  • Traffic-pattern diversification
  • Transport camouflage
  • DPI resistance mechanisms
  • Configurable obfuscation levels

Obfuscation profiles should provide configurable tradeoffs between privacy, performance, latency, and resource consumption.

Core DNS Privacy Module

The DNS Privacy Module protects domain resolution.

Capabilities include:

  • Encrypted DNS
  • DNS over HTTPS
  • DNS over QUIC
  • DNS leak prevention
  • Configurable resolvers
  • Per-profile DNS policies
  • Per-channel DNS policies
  • DNS routing through Umbrix nodes
  • Resolver failover
  • Local DNS interception
  • DNS cache controls

Core Kill Switch Module

The Kill Switch Module prevents traffic from bypassing Umbrix.

Capabilities include:

  • Device-wide kill switch
  • Application-specific kill switch
  • AI-agent-specific kill switch
  • Channel-specific kill switch
  • Interface monitoring
  • Route enforcement
  • DNS leak prevention
  • Emergency traffic blocking
  • Fail-closed behavior

When enabled, the kill switch should prevent traffic from escaping outside the configured Umbrix privacy boundary.

Core Privacy Policy Module

The Privacy Policy Module provides centralized policy enforcement.

Policies may control:

  • Required hop count
  • Allowed regions
  • Restricted regions
  • Node requirements
  • Transport requirements
  • DNS requirements
  • Application routing
  • AI-agent routing
  • Privacy level
  • Obfuscation level
  • Failover behavior
  • Logging restrictions
  • Plugin permissions
  • Channel permissions
  • Bandwidth limits
  • Connection destinations

Core Privacy Profile Module

Umbrix should provide configurable privacy profiles.

Standard

Balanced privacy, performance, and resource consumption.

Advanced

Multi-node routing with increased obfuscation and stronger privacy controls.

Maximum

Aggressive privacy controls, increased route diversity, and stronger traffic protection.

AI Agent

Dedicated privacy settings optimized for individual AI-agent channels.

Custom

Fully user-defined privacy policies.

Profiles should communicate their security and performance tradeoffs clearly.

Core MultiChannel Module

The MultiChannel Module provides independent Umbrix network channels for applications, services, and AI agents.

Each channel represents an independent logical privacy boundary.

Capabilities include:

  • Per-application VPN channels
  • Per-agent VPN channels
  • Independent encryption sessions
  • Independent routing policies
  • Independent DNS policies
  • Independent privacy profiles
  • Independent obfuscation profiles
  • Independent kill switches
  • Independent exit-node policies
  • Independent bandwidth policies
  • Independent security policies
  • Independent channel lifecycle
  • Channel health monitoring
  • Channel failover
  • Channel suspension
  • Channel revocation
  • Channel destruction
  • Channel quotas
  • Channel priority
  • Channel authorization
  • Dynamic channel creation
  • Dynamic channel termination

A channel should be capable of operating independently from other channels on the same device.

Core AI Agent Channel Module

The AI Agent Channel Module extends MultiChannel functionality specifically for AI agents.

Each AI agent can receive a dedicated Umbrix channel.

An AI agent channel can include:

  • Dedicated network identity
  • Dedicated session keys
  • Dedicated route
  • Dedicated entry node
  • Dedicated relay nodes
  • Dedicated exit node
  • Dedicated DNS policy
  • Dedicated privacy profile
  • Dedicated obfuscation policy
  • Dedicated kill switch
  • Dedicated bandwidth limit
  • Dedicated destination policy
  • Dedicated security policy
  • Dedicated plugin permissions
  • Dedicated channel lifecycle
  • Dedicated health state

Multiple AI agents can operate simultaneously without requiring them to share the same routing or privacy configuration.

Core Agent Lifecycle Module

The Agent Lifecycle Module manages network privacy throughout the life of an AI agent.

Supported lifecycle states include:

  • Agent registration
  • Channel creation
  • Policy assignment
  • Route establishment
  • Agent startup
  • Active monitoring
  • Route rotation
  • Policy updates
  • Channel suspension
  • Agent termination
  • Channel destruction
  • Ephemeral key destruction

When an agent terminates, its channel and ephemeral session material should be securely disposed of according to configured policy.

Core Agent Authorization Module

The Agent Authorization Module controls which applications and AI agents may access specific Umbrix channels.

Capabilities include:

  • Agent identity registration
  • Application identity registration
  • Channel ownership
  • Channel access control
  • Permission management
  • Token-based authorization
  • Local authorization
  • Channel revocation
  • Permission expiration
  • Policy inheritance controls

An AI agent should not automatically gain access to another agent’s channel.

Core Agent API Module

The Agent API provides local interfaces for AI orchestration systems and application frameworks.

Capabilities include:

  • Create channel
  • Assign channel
  • Query channel
  • Update channel policy
  • Request route
  • Rotate route
  • Suspend channel
  • Resume channel
  • Revoke channel
  • Destroy channel
  • Query health
  • Query connection state
  • Apply privacy profile
  • Apply DNS policy
  • Apply destination policy

The API should authenticate requests and enforce channel-specific authorization.

Core Agent SDK Module

The Agent SDK provides developers with interfaces for integrating Umbrix into AI applications.

Potential integrations include:

  • Locally hosted AI
  • AI agent frameworks
  • Agent orchestration systems
  • Autonomous applications
  • AI development environments
  • Model servers
  • Tool-using agents
  • Workflow engines

The SDK should abstract channel lifecycle operations without exposing unnecessary cryptographic material.

Core Application Routing Module

The Application Routing Module provides network policies for individual applications.

Capabilities include:

  • Full-device routing
  • Application-specific routing
  • Split tunneling
  • Application groups
  • Per-application DNS
  • Per-application exit policies
  • Per-application privacy profiles
  • Application-specific kill switches
  • Application-specific channel assignment

Core Node Module

The Node Module allows infrastructure to participate in the Umbrix network.

Supported node roles include:

  • Entry node
  • Relay node
  • Exit node
  • Peer relay
  • Private relay
  • Organization-controlled node

Node capabilities include:

  • Secure node identity
  • Key management
  • Route participation
  • Health reporting
  • Capacity reporting
  • Policy enforcement
  • Node isolation
  • Secure configuration

Core Node Discovery Module

The Node Discovery Module provides decentralized infrastructure discovery.

Capabilities include:

  • Distributed node discovery
  • Node capability advertisement
  • Cryptographic node identity
  • Transport capability discovery
  • Node availability
  • Node expiration
  • Discovery caching
  • Privacy-preserving node metadata
  • Regional diversity information

Core Reputation Module

The Reputation Module evaluates node reliability and trust characteristics.

Potential inputs include:

  • Availability
  • Reliability
  • Latency
  • Packet loss
  • Protocol compliance
  • Historical stability
  • Security reports
  • Identity continuity
  • Operator policies

Reputation should inform routing without becoming an absolute centralized authority.

Core Security Module

The Security Module provides system-wide defensive capabilities.

Capabilities include:

  • Session integrity monitoring
  • MITM detection
  • Replay detection
  • Route integrity validation
  • Node behavior validation
  • Protocol downgrade protection
  • Configuration validation
  • Security event detection
  • Secure update verification
  • Plugin security enforcement
  • Channel security monitoring
  • Agent security monitoring

Core Self-Healing Module

The Self-Healing Module provides automatic network recovery.

Capabilities include:

  • Failed-node replacement
  • Broken-route recovery
  • Tunnel reconnection
  • Transport switching
  • Session key rotation
  • Unhealthy-node removal
  • Channel recovery
  • Agent-channel recovery
  • Network failure recovery
  • Policy restoration

Core Telemetry Privacy Module

Umbrix should minimize unnecessary telemetry.

Capabilities include:

  • Local-first diagnostics
  • Opt-in metrics
  • Metadata minimization
  • Anonymous performance statistics
  • Configurable diagnostic levels
  • Local diagnostic storage
  • Secure diagnostic deletion

User traffic contents should not be collected as ordinary telemetry.

Core Configuration Module

The Configuration Module manages system, channel, agent, node, routing, security, and plugin configuration.

Capabilities include:

  • Configuration validation
  • Policy validation
  • Profile management
  • Configuration versioning
  • Secure configuration storage
  • Configuration import
  • Configuration export
  • Configuration rollback
  • Conflict detection

Optional Plugin Modules

Optional plugins extend Umbrix without requiring specialized functionality in the core installation.

Plugins must use documented interfaces and explicit permission models.

AI Threat Detection Plugin

Provides optional AI-assisted detection of:

  • Traffic anomalies
  • Suspicious node behavior
  • Route anomalies
  • Connection manipulation
  • Network attacks
  • Potential traffic-analysis patterns
  • Unexpected agent behavior

AI recommendations should remain subject to Umbrix security policies.

AI Obfuscation Plugin

Provides adaptive selection of approved traffic-obfuscation strategies based on:

  • Network characteristics
  • Transport compatibility
  • Connection performance
  • Privacy requirements
  • Configured policy

Traffic Morphing Plugin

Provides additional traffic transformation capabilities.

Potential capabilities include:

  • Flow normalization
  • Packet-size profiles
  • Timing profiles
  • Adaptive padding
  • Cover traffic
  • Traffic-pattern variation
  • Transport behavior simulation

Advanced Stealth Transport Plugin

Provides additional transport implementations for environments where conventional VPN traffic experiences interference.

All transports must preserve Umbrix authentication and encryption.

Browser Privacy Plugin

Provides browser-level privacy controls.

Capabilities may include:

  • Fingerprint reduction
  • WebRTC leak protection
  • Browser DNS protection
  • Geolocation permission controls
  • Canvas privacy controls
  • WebGL privacy controls
  • User-agent policies
  • Tracking protection

Browser-level protections should clearly communicate platform limitations.

Geolocation Privacy Plugin

Provides optional application-level location privacy controls.

Capabilities may include:

  • Location permission policies
  • Configurable location behavior
  • Location consistency controls
  • Privacy-preserving location responses
  • Optional location simulation

Hardware Privacy Plugin

Provides supported platform-specific privacy controls for:

  • Network identifiers
  • Wi-Fi identifiers
  • Bluetooth identifiers
  • Device metadata
  • Hardware telemetry

Capabilities depend on operating-system permissions.

IoT Privacy Plugin

Extends Umbrix to IoT environments.

Capabilities include:

  • IoT subnet routing
  • Device-specific tunnels
  • Device isolation
  • DNS protection
  • Metadata minimization
  • Device policies
  • Gateway deployment

Secure AI Gateway Plugin

Provides a dedicated Umbrix privacy gateway for locally hosted AI infrastructure.

Capabilities include:

  • Model-server traffic protection
  • AI API isolation
  • Local inference isolation
  • Model-server communication protection
  • AI application-specific tunnels
  • AI workload network policies
  • Metadata minimization
  • Agent channel provisioning

Agent Orchestration Plugin

Integrates Umbrix with external AI orchestration systems.

Capabilities may include:

  • Automatic agent registration
  • Automatic channel creation
  • Agent-to-channel assignment
  • Dynamic policy assignment
  • Agent shutdown detection
  • Channel cleanup
  • Channel health integration
  • Agent fleet management

Container Network Plugin

Provides network isolation for containerized AI agents and applications.

Capabilities include:

  • Container-specific channels
  • Namespace-aware routing
  • Per-container DNS
  • Container kill switches
  • Container-specific exit policies
  • Container channel lifecycle
  • Container network isolation

Virtual Machine Network Plugin

Provides independent Umbrix channels for virtual machines.

Capabilities include:

  • VM-specific channels
  • VM routing policies
  • VM DNS policies
  • VM kill switches
  • VM-specific privacy profiles
  • VM channel lifecycle management

Nested Tunnel Plugin

Provides optional layered tunnel architectures for deployments requiring additional routing separation.

Peer Relay Plugin

Allows authorized clients to operate as temporary relay nodes.

Capabilities include:

  • Ephemeral relays
  • Peer discovery
  • Relay authorization
  • Relay expiration
  • Capacity controls
  • Relay privacy policies

Decentralized Payment Plugin

Provides optional infrastructure for node operators and service providers to coordinate payments without placing payment functionality in the Umbrix core.

Network Simulation Plugin

Provides research and development capabilities for:

  • Simulated network topologies
  • Node behavior testing
  • Routing experiments
  • Traffic-pattern testing
  • Performance benchmarking
  • Failure simulation
  • Protocol testing

Geo Intelligence Plugin

Provides optional network intelligence for:

  • Route recommendations
  • Latency analysis
  • Regional connectivity
  • Node availability
  • Connectivity disruption detection
  • Infrastructure analysis

Node Placement Plugin

Assists infrastructure operators with node placement analysis based on:

  • Network latency
  • Regional diversity
  • Reliability
  • Capacity
  • Route diversity
  • Infrastructure availability

Secure Application SDK Plugin

Provides APIs for integrating Umbrix privacy capabilities into third-party applications.

Potential integrations include:

  • Messaging
  • VoIP
  • AI applications
  • Enterprise applications
  • IoT software
  • Privacy-focused browsers
  • Developer tools

Experimental Cryptography Plugin

Provides a controlled environment for evaluating emerging cryptographic algorithms and post-quantum mechanisms.

Experimental algorithms must not silently replace production cryptographic mechanisms.

Multi-Agent Privacy Architecture

Umbrix should support an environment where multiple AI agents operate simultaneously while maintaining independent network privacy boundaries.

A deployment may contain:

  • Agent A with a two-hop route
  • Agent B with a three-hop route
  • Agent C with a dedicated exit policy
  • Agent D with restricted destination access
  • Agent E with maximum obfuscation

Each agent can maintain independent:

  • Channels
  • Keys
  • Routes
  • DNS policies
  • Privacy profiles
  • Kill switches
  • Security policies
  • Bandwidth limits
  • Plugin permissions
  • Destination policies

The system should prevent accidental channel sharing unless explicitly authorized.

AI Agent Channel Isolation

Each AI agent should be treated as an independently addressable network workload.

A channel should provide:

  • Network isolation
  • Cryptographic isolation
  • Routing isolation
  • DNS isolation
  • Policy isolation
  • Security isolation
  • Lifecycle isolation

The channel architecture should support local AI systems running multiple simultaneous agents without requiring all agents to share one VPN identity or route.

AI Agent Network Lifecycle

A typical AI agent lifecycle should support:

Register → Authenticate → Create Channel → Assign Policy → Establish Route → Start Agent → Monitor → Rotate Route → Update Policy → Terminate Agent → Destroy Channel

Channel destruction should invalidate the associated session and dispose of ephemeral cryptographic material according to configured security policy.

Control Plane

The Umbrix control plane manages distributed network coordination.

Capabilities include:

  • Node discovery
  • Node capability advertisement
  • Route coordination
  • Policy distribution
  • Network health information
  • Node reputation
  • Version compatibility
  • Channel coordination
  • Agent channel management
  • Key-management coordination

Compromise of a discovery or coordination component should not automatically expose encrypted user traffic.

Routing Architecture

Umbrix supports configurable multi-node paths.

A typical route may consist of:

Client or Agent → Entry Node → Relay Node → Exit Node → Destination

Routing can be independently configured for each device, application, container, virtual machine, or AI agent.

Users may configure:

  • Hop count
  • Geographic requirements
  • Node requirements
  • Performance requirements
  • Route rotation
  • Exit policies
  • Relay diversity
  • Privacy profiles
  • Agent-specific policies

Umbrix should not claim absolute anonymity or guarantee that traffic analysis is impossible.

Protocol Architecture

Umbrix uses a protocol architecture that separates:

  • Authentication
  • Encryption
  • Routing
  • Transport
  • Obfuscation
  • Session management
  • Channel management

The protocol should support multiple transports without coupling the security model to one transport.

Planned capabilities include:

  • Ephemeral keys
  • Forward secrecy
  • Replay protection
  • Session rotation
  • Protocol negotiation
  • Transport negotiation
  • Packet authentication
  • Routing envelopes
  • Cryptographic agility
  • Channel-specific sessions
  • Agent-specific sessions
  • Version compatibility

Privacy Architecture

Umbrix treats privacy as multiple independent layers.

Network Layer

Protects IP addresses, routing information, and network traffic.

Transport Layer

Provides encrypted communication and configurable traffic obfuscation.

DNS Layer

Protects domain-resolution activity.

Application Layer

Provides application-specific routing and privacy policies.

AI Agent Layer

Provides dedicated network channels and privacy boundaries for individual AI agents.

Browser Layer

Provides optional browser fingerprint and privacy controls.

Device Layer

Provides supported hardware and telemetry privacy controls.

Infrastructure Layer

Provides decentralized node discovery, routing, and operator controls.

Security Architecture

Umbrix follows a defense-in-depth model.

Security mechanisms include:

  • Authenticated encryption
  • Forward secrecy
  • Ephemeral keys
  • Key rotation
  • Replay protection
  • Secure random generation
  • Route integrity verification
  • Node authentication
  • Protocol downgrade protection
  • Secure updates
  • Kill switch enforcement
  • Configuration validation
  • Plugin isolation
  • Channel isolation
  • Agent isolation
  • Security auditing

Cryptography

The initial cryptographic architecture should support:

  • X25519
  • ChaCha20-Poly1305
  • BLAKE3
  • Cryptographically secure randomness
  • Ephemeral session keys
  • Per-channel keys
  • Per-hop keys
  • Forward secrecy
  • Hybrid post-quantum key exchange

Umbrix should maintain cryptographic agility so algorithms can evolve without requiring a complete architectural redesign.

Cross Platform Support

Umbrix targets:

  • Windows
  • Linux
  • macOS
  • Android
  • iOS

Optional integrations include:

  • Browser extensions
  • IoT gateways
  • Self-hosted servers
  • Cloud nodes
  • Developer SDKs
  • Container environments
  • Virtual machines
  • Locally hosted AI systems
  • AI orchestration systems

Platform-specific capabilities depend on operating-system permissions and security models.

AI Privacy

Umbrix is designed to provide a privacy layer for locally hosted AI.

Running an AI model locally does not automatically protect:

  • Network traffic
  • DNS requests
  • API connections
  • Remote tools
  • Model downloads
  • Update services
  • External data sources
  • Application telemetry
  • Agent-to-agent communications

Umbrix should allow each AI agent to receive its own network channel so that local AI deployments can establish independent privacy boundaries for every running agent.

The long-term objective is to make network privacy a standard component of locally hosted AI infrastructure.

Developer Platform

Umbrix is designed for extensibility.

Developers can build integrations for:

  • Routing
  • Transport
  • Obfuscation
  • Privacy controls
  • AI security
  • Browser integration
  • IoT networking
  • Application integration
  • Cryptographic experimentation
  • Monitoring
  • Agent orchestration
  • AI development frameworks

Plugins should expose explicit capabilities and permissions.

Plugin Security Model

Every plugin should declare:

  • Plugin name
  • Version
  • Required Umbrix version
  • Permissions
  • Network access requirements
  • Data access requirements
  • Configuration requirements
  • Security dependencies

Plugin permissions should follow least-privilege principles.

Security-critical core functionality must not be replaceable by untrusted plugins without explicit administrator approval.

Configuration

Umbrix configuration should support:

  • Privacy profiles
  • Routing policies
  • Node preferences
  • Hop counts
  • DNS policies
  • Kill switch settings
  • Application routing
  • AI-agent routing
  • Transport preferences
  • Obfuscation profiles
  • Channel policies
  • Agent policies
  • Plugin configuration
  • Update channels
  • Diagnostic preferences

Configuration should be validated before activation.

Deployment

Umbrix should support:

  • Individual clients
  • Self-hosted nodes
  • Private organizational networks
  • Distributed node networks
  • Cloud deployments
  • Hybrid deployments
  • IoT gateways
  • Containerized deployments
  • Virtual machines
  • Locally hosted AI environments
  • AI agent environments

Node operators should be able to configure infrastructure independently without requiring access to user traffic or private client credentials.

Commercial Grade Operations

Umbrix should support production deployments through:

  • Stable releases
  • Beta releases
  • Signed releases
  • Automatic updates
  • Health monitoring
  • Node failover
  • Route recovery
  • Configuration validation
  • Security advisories
  • Audit-friendly architecture
  • Deployment automation
  • Optional metrics
  • Version compatibility checks
  • Channel health monitoring
  • Agent fleet management

Testing and Verification

Umbrix development should include:

  • Unit testing
  • Integration testing
  • Protocol testing
  • Cross-platform testing
  • Multi-node testing
  • Multi-channel testing
  • Multi-agent testing
  • Failure testing
  • Performance testing
  • Privacy leak testing
  • DNS leak testing
  • Route integrity testing
  • Cryptographic testing
  • Fuzz testing
  • Plugin isolation testing
  • Channel isolation testing
  • Agent isolation testing
  • Security auditing

Privacy claims should be validated through measurable testing.

Performance

Umbrix should provide configurable tradeoffs between:

  • Privacy
  • Latency
  • Throughput
  • Battery consumption
  • CPU usage
  • Route complexity
  • Obfuscation strength
  • Number of channels
  • Number of active AI agents

Users should be able to select appropriate profiles rather than forcing maximum overhead on every connection.

Observability

Umbrix should provide privacy-preserving operational visibility.

Supported information may include:

  • Connection status
  • Tunnel health
  • Channel health
  • Agent channel status
  • Node availability
  • Latency
  • Packet loss
  • Route state
  • Protocol state
  • Resource usage

Sensitive user traffic contents should not be included in ordinary diagnostics.

Emergency Controls

Umbrix should provide emergency controls capable of:

  • Blocking network traffic
  • Terminating active tunnels
  • Terminating individual channels
  • Terminating individual agent channels
  • Removing ephemeral routing information
  • Rotating session keys
  • Disconnecting compromised nodes
  • Disabling plugins
  • Revoking agent permissions
  • Restoring safe configuration defaults

Emergency controls should prioritize preventing accidental traffic leakage.

Security Disclosure

Security vulnerabilities should be reported responsibly through the project’s designated security-reporting process.

Security researchers should avoid publicly disclosing exploitable vulnerabilities before maintainers have had an opportunity to evaluate and address them.

Specification Branding License (SBL)

Standard

Optional

License & Notice Requirements

Umbrix is released under the GNU Affero General Public License v3.0 or later (AGPL-3.0+).
By contributing to this project, you agree that your contributions will also be released under this license.

Please note the following:

  • All contributions must comply with the AGPL-3.0+ terms.
  • Under Section 7 of the license, all redistributions, forks, and derivative works must preserve attribution to:
    Roxanne Ardary and roxanneardary.com.
  • Umbrix specifications are free to use with attribution. A Specification Branding License can be negotiated upon request.
  • The project’s notice.md file tracks attribution requirements and contributor acknowledgments. Any update that adds new contributors or modifies attribution should also update notice.md.
  • When submitting a pull request, ensure that any new files maintain the attribution headers where applicable.
  • Network-deployed versions of this software must also remain fully AGPL-3.0+ compliant, including exposure of source code modifications when applicable under the license.

For full legal details, please refer to the AGPL-3.0+ license and the project’s notice.md file.

Notice – Umbrix

Attribution Requirement: Under Section 7 of the AGPL 3.0+ license, all redistributions, forks, and derivative works, including network-deployed versions of this project, must provide attribution to Roxanne Ardary and roxanneardary.com.

Contributors

This file tracks contributors and their specific contributions to the project.

  • Roxanne Ardary, roxanneardary.com – March 31, 2026
    Created the repository for Umbrix. Defined the core architecture, feature set, and foundational design for a multi-node, fully obfuscated VPN system with AI-assisted adaptive privacy and cross-platform support.
  • [Add other contributors here] – [Date]
    [Describe contribution in one sentence]

License – Umbrix

This repository is licensed under the GNU Affero General Public License v3.0 or later (AGPL-3.0+).

Key Points

  • You are free to use, modify, and distribute the code.
  • All redistributions, forks, and derivative works or network-deployed versions must also be licensed under AGPL-3.0+ and provide attribution to Roxanne Ardary and roxanneardary.com as required under Section 7 of the license.
  • The software is provided “as is,” without warranty of any kind.

For the full license text, see GNU AGPL-3.0 License.

Related posts:

IntegrityLayer Specification

ShellScan

Blockchain