Architect Genesis CLI
A CLI that scores project architecture 0 to 100 from the AST, finds anti-patterns and runs AI refactoring on request.
- 56 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- `architect execute` edits project files through a model, and its requests go to the provider.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/architect-genesisDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- `architect execute` edits project files through a model, and its requests go to the provider.
This is third-party code. Review the repository files before installing.
What it does
Architect Genesis parses code through Tree-Sitter, builds a dependency graph and scores architecture from 0 to 100. TypeScript, Python, Go, Java, Rust, Ruby and PHP are supported. The tool detects layers, stack and domain on its own. The weights of the four score dimensions adapt to the stack. Main commands: `architect analyze ./src` for analysis, `architect check ./src` to validate rules from `.architect.rules.yml`, `architect forecast ./src` to predict decay from git history and `architect execute ./src` for refactoring with Claude, GPT or Gemini. A separate MCP server exposes the same capabilities to Claude Code and Cursor through nine tools.
Who it is for. Tech leads and developers who track architecture health and want the check wired into agents and CI.
Good fit when
- A dependency graph and anti-pattern list are needed before refactoring.
- A score drop below the threshold in `.architect.rules.yml` must fail the CI check.
- The team wants to know from git history where architecture is decaying.
- An agent in Claude Code asks for a score and refactoring hints through MCP.
Not a fit when
- The project is written in a language outside the list of seven.
- Automated AI refactoring without manual review is unacceptable.
- Code security analysis is needed: the tool scores structure.
Example request
Analyze `./src` with architect, show the score, anti-patterns and the three most useful refactoring steps.Limitations
The CLI needs Node.js 18 or newer. `architect execute` changes files with a model, so you need a key for the chosen provider and a review of the result. The score depends on the project and the weights the tool picks for the stack, and is no absolute truth. The `kb` knowledge base stores score history, and the forecast uses linear regression. The README advertises the same author's paid dev team, while the tool itself is open under MIT. The repository holds many service notes and release templates.
How to disable. Uninstall the package with `npm uninstall -g @girardelli/architect` and, if you added MCP, remove its entry from the client configuration.
FAQ
How do I install Architect and run an analysis?
Run `npm install -g @girardelli/architect`, then `architect analyze ./src`. No setup is needed: the tool detects stack and domain itself.
How do I connect the MCP server?
Add the `npx @girardelli/architect-mcp` launch to the client configuration. It opens nine tools, including `analyze_project`, `get_score` and `suggest_refactoring`.
How do I define rules for the check?
Describe them in `.architect.rules.yml`, for example the minimum overall score `min_overall_score`. `architect check` compares the project with those rules.
What does `architect forecast` do?
It reads git history and predicts with linear regression how the score will keep falling. Results go to the knowledge base, and `architect kb history .` shows the trend.
Which languages are supported?
TypeScript, Python, Go, Java, Rust, Ruby and PHP. Parsing goes through Tree-Sitter.
Discussion
No questions or reviews yet.
Related
CLI toolsArchgate turns architecture decisions into executable rules and blocks violations in CI, hooks and AI agent work
A Claude Code plugin for project architecture: ADRs, multi-role reviews, findings turned into plans, a pragmatic mode.
A Claude Code plugin that interviews you, designs the architecture and writes a self-contained blueprint to build from
Git-native project context for agents: specs, decisions and rules live in Markdown beside the code and are read before edits
An architecture coach skill: questions lead to ADRs, a diagram and a plan, and a repository is audited from its code
CodeGraph (Rust)
codegraph-rust
A Rust code graph for agents: four MCP tools show what an edit will touch, how the architecture fits together and where the fragile spots are
README in short
The tool builds a dependency graph, assesses layers and dimension weights by stack, and a knowledge base keeps score history. Four agents exist: review, forecast, refactor and module scaffold. A VS Code extension shows scores in CodeLens.