simulate.farm
PlatformDataScienceDocumentationAbout
Log inRequest access
System Architecture

MCP Server

This page describes the planned Model Context Protocol (MCP) coordination layer. Public MCP access, OAuth, job scheduling, and production simulation execution are not yet available.

Current status: architecture preview. The first production slice is establishing authenticated REST scenario materialization and a pinned RuFaS runtime contract before MCP is exposed.

Core Capabilities

These are target capabilities and should not be read as claims about the currently deployed product.

Request Interpretation

Convert conversational requests into valid simulation parameters. The server parses natural language inputs and maps them to RuFaS configuration schemas with automatic validation.

Execution Coordination

Manage the setup and execution of RuFaS simulations. The server handles job queuing, resource allocation, and status monitoring throughout the simulation lifecycle.

Controlled Data Flow

Ensure that simulation inputs and outputs move safely between system components. All data transfers are logged, validated, and encrypted in transit.

Secure System Integration

Provide a controlled bridge between user interactions and the simulation engine. The server enforces access controls and prevents unauthorized operations.

System Architecture

The proposed MCP server will sit between authorized clients and shared platform application services; it will not bypass their ownership, region, or mutation controls.

Interface Layer
Conversational UIWeb DashboardAPI Clients

User-facing components that send requests to the MCP server

MCP Server
Request ParserValidatorJob SchedulerResult Handler

Central coordination layer managing all simulation operations

Execution Layer
RuFaS EngineCompute ResourcesStorage Backend

Backend systems that perform simulations and store results

Communication Protocols

The planned MCP interface will use structured protocols backed by the same authorization and application-service boundaries as REST.

Request Protocol

Structured JSON messages containing simulation parameters, user context, and execution preferences.

Validation Protocol

Parameter range checking, schema validation, and consistency verification before execution.

Execution Protocol

Job submission, progress tracking, and completion notification with error handling.

Response Protocol

Structured result delivery with metadata, provenance information, and export options.

Technical Implementation

The intended implementation uses an asynchronous job boundary and idempotent application services. It has not yet been deployed or load-tested.

The current slice introduces allowlisted audit records and immutable input manifests. Complete MCP and execution audit coverage remains future work.

  • Stateless request processing for horizontal scalability
  • Idempotent operations for reliable retry behavior
  • Comprehensive error handling with informative messages
  • Rate limiting and resource quotas per user/project
Processing

Asynchronous job queue with priority scheduling and automatic retry

State Management

Persistent job state with checkpointing for long-running simulations

Security

Token-based authentication with role-based access control

Integrate with the MCP Server

Learn how to connect your applications to the simulation infrastructure through our enterprise APIs.

Explore Enterprise APIsView Documentation
simulate.farm

Scientific farm simulation that keeps configuration, assumptions, runs, and results connected.

Product

Interactive demoPlatformDocumentation

Company

AboutData & securityPasalica

© 2026 Pasalica. All rights reserved.

RuFaS is the first model integration.