jgs-sysmlv2-api-mcp

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.

Category
访问服务器

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:

  1. The commit POST body is a CommitRequest, not a Commit. There is no client-supplied top-level @id; the server assigns the commit id.
  2. The official client's typed Element model drops declaredName and other SysML payload fields. Reads go through a raw-JSON path instead of typed models.
  3. 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 previousCommit parentage, 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.

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选