Build and maintain the model
Turn requirements and design documents into packages, elements, connectors and diagrams in your repository, then prove the result by counting back and rendering it.
Build a model from source documents →Free tools that put your AI assistant to work where architects actually work: in your live Sparx Enterprise Architect repository and in your Visio diagrams. Ask in plain language; the assistant builds, checks and reports from the real model, and you keep the decisions that need an architect.
The tools follow the four use cases of AI-Augmented Enterprise Architecture. Each comes with step-by-step AI Augmentation Scenarios: what to ask for, what you end up with, and how to check it.
Turn requirements and design documents into packages, elements, connectors and diagrams in your repository, then prove the result by counting back and rendering it.
Build a model from source documents →What depends on this system before you retire it? How does a requirement reach the systems that realize it? What changed, when, and by whom? Answers come from the live model, with a plain statement of how far the search went.
Trace dependencies →Run your modeling rules against a model or a single change, catch what does not conform before it is promoted, and keep your modeling language under control. The tools confirm completeness and conformance; judging whether the architecture is right stays with you.
Catch a non-conforming change →Turn repository data into portfolio answers, a reporting database or a Power BI model, in the language of the people who asked, without teaching them Enterprise Architect.
Answer a stakeholder portfolio question →Both connect your AI assistant to the application you already run — the live model or the live document, not a copy or an export.
Documentation for the Visio tools is on the way.
Pick the AI client you work in. Both are free, run on Windows alongside Enterprise Architect, and need no registration or license key. Free to use under the AI Power Tools license agreement.
For Claude Code, Claude Cowork and Claude Desktop chat. Two pieces, each installed once.
1. The extension: the AI Power Tools MCP Server for Sparx EA
Drag the downloaded file onto Claude Desktop’s Extensions pane, then click Install. Claude Code, Claude Cowork and Claude Desktop chat all use this server.
Download the extension (.mcpb)2. The plugin: the skills
In Claude Desktop or Claude Cowork, add the novocircle marketplace at claude.ai/customize/plugins and install ai-power-tools, or upload the plugin file. In Claude Code:
claude plugin marketplace add NovoCircle/ai-power-tools-skills
claude plugin install ai-power-tools@novocircle
For VS Code with GitHub Copilot in Agent mode (VS Code 1.99 or later). Copilot gets the AI Power Tools MCP Server for Sparx EA and works directly in your Enterprise Architect repository; the scenarios show what to ask for.
In VS Code:
Extensions view → … menu → Install from VSIX…, then choose the downloaded file. Or from a terminal: code --install-extension ai-power-tools-ea.vsix
Keep the extension and the skills on the same release: the server tells you when either one is behind, and how to update it. Not sure which to choose, or deploying to a fleet? See which package is right for you.
Free to use under the AI Power Tools license agreement.
Windows 10 or 11, on the same machine as the application the tools work in.
Sparx Systems Enterprise Architect 16.1 or later, installed and licensed. Or Microsoft Visio, for the Visio tools.
Claude Code, Claude Cowork or Claude Desktop chat; or VS Code 1.99 or later with GitHub Copilot in Agent mode. What each can do is in capabilities and surfaces.
The tools and skills don’t update themselves. The server tells you when a newer release is out. Sign up to get an email too, along with other resources for getting the most from AI Power Tools. The downloads above work whether or not you sign up.
The tools do the repeatable work. Getting value from them is about the practice around them — and we can help with that too.
The practice: start with the real problem, then put AI to work on modeling, analysis, governance and stakeholder engagement.
See the role specialty →Core AI foundations for your architects, then architect-specific skills — as a cohort or one-on-one on your own architecture work.
See 1:1 mentoring →Our dedicated team of Sparx Enterprise Architect specialists, bringing AI-augmented tooling, methodology and hands-on delivery into your EA environment.
Visit Sparx Services →Questions about getting your team set up, or want to talk through what AI-augmented architecture could look like for you?
Book a Discovery Call