jgs-sysmlv2-api-mcp
An MCP server that gives an AI agent read and write access to a live SysML v2 model through the vendor-neutral OMG SysML v2 REST API.
README
<!-- Copyright (c) 2026 JG Systems Consulting Ltd. See LICENSE. -->
jgs-sysmlv2-api-mcp
An MCP server that gives an AI agent read and write access to a live SysML v2 model through the vendor-neutral OMG SysML v2 REST API.
What it is, who it's for
This server talks to any conformant SysML v2 API endpoint: the OMG reference pilot, or a commercial implementation. There is no CATIA Magic, no desktop tool, no GUI in the loop. If your systems model already lives in a headless SysML v2 API repository, and you want an AI agent to explore it, extend it, or restructure it directly, this is the bridge.
Writes go through a staged-commit model. create, modify, and delete operations accumulate in an in-process buffer, get reviewed as a batch, and are posted as one atomic commit gated by a confirmation token bound to the exact buffer contents. Nothing reaches the API until you have seen what is about to be sent.
It is for teams and individual engineers who want an agent working directly against a SysML v2 API project: exploring an existing model, authoring new elements and relationships, running parallel branches per agent session, or managing project lifecycle, all through the standard API rather than a proprietary desktop integration.
Companion to
jgs-magic-sysmlv2-mcp, which drives a running Cameo instance instead. Use this server when the model lives in a headless SysML v2 API repository; use the Cameo bridge when it lives in a running Cameo desktop session.
Install
pip install -e .
pip install "git+https://github.com/Systems-Modeling/SysML-v2-API-Python-Client.git"
Requires Python 3.11 or later. See docs/install.md for prerequisites, the local pilot Docker recipe, and how to verify the install with the ping tool.
Usage
Set the required environment variables and run the server over stdio:
export SYSMLV2_BASE_URL=https://your-sysmlv2-endpoint.example.com
export SYSMLV2_TOKEN=your-bearer-token
export SYSMLV2_PROJECT=your-project-uuid
python -m jgs_sysmlv2_api_mcp
The server exposes 35 tools across five areas: navigation reads, staged authoring writes, project lifecycle, branch and tag versioning, and infrastructure (ping, stats). See docs/usage.md for four fully walked workflows (exploring an unfamiliar model, authoring and committing a change, running one branch per agent session, and managing project lifecycle), and docs/TOOL-REFERENCE.md (or the browsable docs/index.html / docs/tool-reference.html pages) for the complete, generated tool list with parameters.
Install with your AI agent
Repository: https://github.com/jgsystemsconsulting/jgs-sysmlv2-api-mcp
Version: v1.0.1
Install:
pip install -e .
pip install "git+https://github.com/Systems-Modeling/SysML-v2-API-Python-Client.git"
Human step (cannot be automated): add this server to your MCP client's
config (see examples/.mcp.json.example) and restart the client so it
picks up the new server.
This repo also ships host-native plugin manifests, so it can be discovered from inside a
coding-agent host rather than hand-edited into an MCP config: Claude Code
(/plugin marketplace add jgsystemsconsulting/jgs-sysmlv2-api-mcp), Cursor
(.cursor-plugin/, point Cursor at this repo), OpenAI Codex CLI (.agents/plugins/marketplace.json,
also read via the legacy .claude-plugin/ path), and Gemini CLI
(gemini extensions install https://github.com/jgsystemsconsulting/jgs-sysmlv2-api-mcp).
License
Proprietary, copyright 2026 JG Systems Consulting Ltd. See LICENSE for full terms and docs/licensing.md for a plain-language summary and how to request a license.
Support
Open an issue on GitHub for bugs and questions. For security vulnerabilities, follow the private reporting process in SECURITY.md rather than opening a public issue.
Design notes
Three ground-truth facts, discovered by a day-one spike against the pilot, shape the code:
- The commit POST body is a
CommitRequest, not aCommit. There is no client-supplied top-level@id; the server assigns the commit id. - The official client's typed
Elementmodel dropsdeclaredNameand other SysML payload fields. Reads go through a raw-JSON path instead of typed models. - Reconciliation keys on branch head-advance, not a client-minted commit id. The commit list is not guaranteed to be newest-first, so head is always resolved by walking
previousCommitparentage, never by trusting list order.
The full spike results, design decisions, and tier-by-tier implementation history live in docs/dev/, which is engineering history, not customer documentation.
Version
Current release: v1.0.1. See CHANGELOG.md for what changed.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.