Runtime platform

ULTRAMEGA S1

Run and monitor your industrial software modules.
ULTRAMEGA S1 manages module processes, communication, configuration and recovery. Connect your own software through .NET, gRPC or REST, and use the built-in MCP endpoint to give AI tools controlled access.

Control C# / .NET Analytics Python Integration C++ / gRPC AI Agent MCP ... S1 Gateway REST API · WebSocket Events · SignalR Hub · Auth · RBAC S1 Platform (Hosting) Process Lifecycle · Port Allocation · Self-Healing · Health Monitoring Fault Isolation Separate processes Real-time IPC Low-latency messaging Open APIs .NET · gRPC · REST Self-Healing Three-layer resilience Core NuGet packages: Client · Hosting · Runtime.{rid}
Overview

What is ULTRAMEGA S1?

ULTRAMEGA S1 is a runtime platform for engineering teams building modular test, measurement and automation systems. It starts and monitors module processes, routes messages and provides APIs for configuration and diagnostics. Your team develops the application logic; S1 provides the shared runtime services.

Use S1 with independently developed modules or as the runtime for S1 Foundry. Built-in services support multi-step workflows, scheduled execution, message history and communication between S1 instances. Configuration management tracks the difference between the requested settings and those applied by each module. The developer package and NuGet packages provide the tools and binaries needed to integrate S1 into your application.

MODULES AND CLIENTS — .NET · gRPC · REST · MCP DAQ Adapter Adapter worker Data Flow Python PLC Bridge C++ · gRPC AI Agent MCP Alarm Manager C# · .NET ULTRAMEGA S1 RUNTIME PLATFORM Process isolation · Message bus · Health & recovery · REST · WebSocket · MCP Gateway Instruments PLCs · Controllers Historians HMI · SCADA AI services YOUR EXISTING SYSTEMS, HARDWARE & DATA
Core capabilities

Runtime capabilities

Process isolation

Each module runs in a separate process and memory space. The hosting layer manages startup, ports and automatic restarts. This isolates process failures; your application still needs to handle unavailable modules and interrupted work.

Details

Real-time messaging

Exchange events and commands through a low-latency inter-process message bus. Use fire-and-forget messages for events and synchronous commands for request-response operations, with ordered delivery where configured. Typed messages and schema validation help detect incompatible data at module interfaces.

Details

REST API & event streaming

Manage modules, send messages and commands, and inspect configuration, health and diagnostics through 11 HTTP API groups. Receive events over WebSocket or use SignalR for module-specific message delivery.

Details

Self-healing architecture

Circuit breakers stop repeated calls to unresponsive modules, connection handling retries interrupted gRPC links, and process supervision restarts crashed modules. These mechanisms address different failure modes. State recovery uses saved checkpoints where the module supports them.

Details

Configuration reconciliation

Apply configuration through a single managed path and compare requested settings with the settings reported by the running module. Instance snapshots and configuration fingerprints show whether the change has taken effect without exposing secret values.

Details

Health & observability

Check liveness, readiness, startup status, detailed health and aggregate health. Connect the relevant endpoints to orchestrators and load balancers, and use metrics, audit trails and per-module status to investigate operational problems.

Details

Workflow coordination

Define how work moves between modules using typed messages, routing rules, scheduled delivery and checkpoints. Use sagas for multi-step operations that need defined recovery actions when a step fails.

Details

Saga orchestration

Track progress through a multi-step workflow, limit concurrent execution and identify each operation with a unique correlation ID. If a step fails, registered compensation handlers run the recovery actions defined by your application. These actions are not a general rollback of physical processes.

Details

Federation

Connect multiple S1 instances into a federated mesh. Messages route transparently across instance boundaries with endpoint validation and SSRF protection. Federation health monitoring tracks peer availability and enforces forwarding timeouts.

Details

Scheduled execution

One-time and recurring scheduled message delivery with persistent storage. Define managed jobs that survive restarts. The scheduler integrates with the feature flag system for runtime kill-switches and supports bounded query pagination across all schedule stores.

Details

Content routing

Route messages based on headers, message type, or destination prefixes. Define routing rules through the API that match message metadata and direct them to the appropriate modules. Add, remove, and update rules at runtime through the API without module restarts.

Details

Journal & checkpoint

Persistent message journal with query and replay capabilities. Save and restore module state through the checkpoint API. Journal compaction runs automatically to manage storage growth, with structured metrics for monitoring retention and drop rates.

Details
Resilience

Self-healing architecture

Connection retries, circuit breakers and process supervision provide distinct ways to detect and recover from failures.

Circuit breakers

When a downstream module stops responding, the circuit opens to prevent request pileup and cascading timeouts. Calls fail fast with a clear status while the breaker periodically probes for recovery.

  • Protects inter-module communication
  • Configurable thresholds per module pair
  • State visible through health API
Details

Resilient connections

Interrupted gRPC connections are retried automatically. Increasing delays with random variation prevent clients from all reconnecting at once. Message delivery behaviour depends on the configured delivery and recovery path.

  • Protects gRPC and TCP transport links
  • Prevents reconnect storms
  • Connection state reported in real time
Details

Process supervision

The hosting layer detects crashed processes, cleans up stale state, and restarts modules with their last known configuration. PID tracking and orphan detection prevent zombie processes from consuming resources.

  • Protects module process lifecycle
  • Rate-limited restarts
  • Checkpoint recovery on restart
Details
AI integration

Model Context Protocol (MCP)

Give AI tools access to S1 operations through explicitly assigned permissions.

MCP endpoint

The built-in MCP server provides 56 Gateway system tools plus tools registered by modules. Authorised clients can manage modules, send messages and inspect health. Access is controlled per session.

Details

Capability-based access control

Assign capabilities such as system:read, system:module_management, messaging:send and system:admin to each session. Role-based access control determines which tools the session can discover and invoke.

Details

Session management & audit

Sessions are associated with an identity, and tool calls are recorded in an audit log. Session broadcasting supports coordination between clients connected to the same S1 instance.

Details

Tool schema registry

Describe custom tools using JSON Schema. Clients can discover the available operations and their parameters through the registry rather than relying on a separately maintained list.

Details
MCP tools

56 Gateway system tools

Gateway system tools and module-defined tools are available to AI agents and human operators through a single interface

Module management

Register, start, stop, restart, and remove modules. Query state and health.

Details

Messaging

Send, batch, broadcast, and query-with-response. Typed messages with routing.

Details

Configuration

Read and update substrate config. Inspect and toggle feature flags through authorized tool access.

Details

System control

Shutdown, restart, system info, performance stats, and platform metadata.

Details

Cluster

Node management, leader election, service discovery, and distributed state.

Details

Gateway info

Read version, about, runtime status, and performance metadata.

Details

Service discovery

Register service endpoints and resolve them by name.

Details

Type system

Register message types with schemas and query the type registry.

Details

Content routing

Add, remove, and list content-based routing rules with predicate matching.

Details

Journal

Query message history, replay sequences, and inspect journal statistics.

Details

Checkpoint

Save, restore, and list module state checkpoints for crash recovery.

Details

Scheduling

Schedule messages for future or recurring delivery. Cancel and list jobs.

Details

Sagas

Start durable multi-step workflows with compensation. Track and cancel.

Details

Federation

Register remote substrate peers, manage federation links, and check peer health.

Details
Distribution

Add S1 to your application

Use the .NET client and hosting packages for integration, and the developer package for SDK tools, templates and runtime binaries.

Ultramega.S1.Client

Typed .NET client for the Gateway REST API. Covers authentication, module management, messaging, synchronous commands, discovery, config, health, metrics, and feature flags. Includes WebSocket event streaming and SignalR hub integration with auto-reconnect.

  • AddS1GatewayClient() for DI
  • AddS1EventStream() for WebSocket
  • AddS1HubClient() for SignalR

Ultramega.S1.Hosting

Managed process hosting for the S1 platform. Spawns and monitors Gateway and ModuleHost processes, handles port allocation, readiness detection, graceful shutdown, and crash recovery. Two modes: managed (owns processes) or external (connects to existing).

  • AddS1Platform() as IHostedService
  • Module registration API
  • Self-healing process supervision with circuit breakers

Developer package & runtime binaries

The developer package contains pre-built platform binaries and SDK files. Runtime packages copy the required binaries into your application's output directory.

  • Developer package artifacts for Windows, Linux, and macOS
  • Zero-dependency self-contained executables
  • Drop-in deployment with no separate runtime prerequisite
Integration

Module integration

Native and gRPC integration

Native modules (C# / .NET)

Implement modules with the .NET SDK to use S1 messaging, lifecycle management and configuration handling. Register them through the hosting API and declare the capabilities they expose to clients.

  • Direct SDK integration via IModule
  • Platform-owned module instance identity
  • Secret-safe configuration fingerprints
  • Capability declaration for MCP/tool access

External modules and clients

Connect existing software through the published APIs and gRPC contracts. External-module v1 supports configuration updates through ApplyConfiguration. Adapter-worker v1 does not support applying configuration in the worker.

  • Python and C++ gRPC external-module paths
  • Wire-level Gateway clients and adapter-worker templates
  • REST API for registration, discovery, and configuration transport
  • Clear contracts for module lifecycle, messaging, and configuration
Applications

Use cases

Examples of where a shared module runtime can be useful.

Industrial control systems

Connect control software from different vendors through modules and adapters. Coordinate Python, C++ and .NET components while managing their processes and communication through S1.

Test & measurement

High-rate test workflows with real-time data acquisition. Coordinate test fixtures, vision systems, and quality databases through sagas, typed messages, routing rules, and scheduled delivery.

System modernisation

Add API connections to existing PLC and SCADA systems for analysis and remote monitoring. Introduce the integration in stages, with testing and commissioning planned around the operating installation.

HIL & digital twins

Connect physical equipment and simulated components for hardware-in-the-loop testing. Use event streams to inspect their behaviour, and verify the timing requirements for the intended test setup.

Robotics & automation

Exchange messages between modules for robot coordination, sensor processing and path planning. Your application provides the algorithms and must meet the timing and safety requirements of the machinery.

AI-assisted operations

Connect AI tools through the built-in MCP endpoint to inspect system state and assist with diagnostics. Assign write and control permissions explicitly, according to your operational procedures.

REST integration

Gateway API

11 endpoint groups for authentication, module management, messaging and operational data.

Authentication

JWT login, token refresh, logout, password change, and token verification.

Details

Modules

Full lifecycle: register, start, stop, restart, delete. Plus health, logs, and external module support.

Details

Messages

Send single, typed, or batch messages. Query delivery status by message ID.

Details

Discovery

Register and resolve services. Load-balanced selection with round-robin strategy.

Details

Health

Liveness, readiness, and detailed health endpoints for load balancers and orchestrators.

Details

Configuration

Read substrate config and apply module configuration with schema/snapshot reads, correlation fields, retry hints, and secret-safe fingerprints.

Details

Features

List all feature flags and check individual flags. Runtime mutation is handled through authorized admin/tool surfaces.

Details

Metrics

JSON and Prometheus formats. Drain diagnostics and runtime metrics.

Details

WebSocket

Real-time event and message streams. Connection info endpoint for clients.

Details

Types

Register and query message types, module states, and delivery guarantee enums.

Details

Diagnostics

System info and drain progress for debugging message processing bottlenecks.

Details
Resources

Documentation

Documentation and packages

Technical whitepaper

Architecture overview, performance characteristics, and the design decisions behind the message-driven microkernel.

Go to Downloads

SDK reference

APIs, data contracts, lifecycle hooks, threading model, resilient connections, and code samples for native module development.

Go to Downloads

Hosting packages guide

Step-by-step adoption guide for S1.Client, S1.Hosting, and S1.Runtime NuGet packages. Includes migration checklist and DI patterns.

Go to Downloads

MCP integration guide

Connecting AI agents to S1. Tool registration, capability scanning, RBAC configuration, and session management.

Go to Downloads

Integration examples

Practical walkthroughs for PLCs, SCADA, enterprise systems, and external module integration via gRPC and WebSocket.

Go to Downloads

Gateway API reference

Complete REST API documentation for the Gateway endpoint groups. Request/response schemas, authentication, and WebSocket protocols.

Go to Downloads

Download ULTRAMEGA S1

The S1 core platform is free, including for commercial use. Download the developer package or use the NuGet packages in your .NET application. Contact us if you need help with integration or system development.