c3-skill

🥈Silver

C3 turns your codebase architecture into a queryable, machine-verifiable markdown structure that LLMs can safely read and update through Claude Code or CLI.

1000Updated 2mo ago

Intermediate30min to implementautomation

Saves ~300 min per use

Quick InstallView Source

git clone https://github.com/lagz0ne/c3-skill.git

Works with:

Claude

Overview

About This Skill

C3 solves architecture documentation drift by making your codebase structure machine-readable and machine-verifiable. It creates a sealed `.c3/` directory of markdown files that define your entity model—containers, components, references, rules, and ADRs—with validation enforced by your team's own definitions. Agents query by meaning using semantic search, look up which component owns a file, trace transitive impact across your architecture, and write changes that pass structural and semantic validation. Use it through the Claude Code plugin with `/c3 onboard this project`, or via the `npx` CLI for batch operations. Git tracks all architectural decisions in reviewable diffs, while `c3.db` remains a local, disposable cache.

How to Use

Install via Claude Code with `claude plugin install lagz0ne/c3-skill`, then run `/c3 onboard this project` to scaffold your `.c3/` directory. Alternatively, use the npm CLI with `npx @c3x/cli search "your question"` or `npx @c3x/cli lookup <file-path>`. Edit canvases in `.c3/canvases/` to define your entity model, and all reads and writes validate against your definitions.

Use Cases

Query codebase architecture by meaning without knowing exact file paths or entity names

Validate that architecture documentation matches your team's definitions before agents modify code

Trace transitive impact of changes across containers, components, and cross-cutting references

Create and enforce architectural standards (ADRs, rules, refs) with machine-readable validation

Setup & Installation

Quick Install

No install command available. Check the GitHub repository for manual installation instructions.

Alternative Install (Git Clone)

git clone https://github.com/lagz0ne/c3-skill

Requirements

Claude Code or compatible AI agent
Works with: Claude

Quick Start Guide

Install the Skill

Copy the install command above and run it in your terminal.

Open Your AI Agent

Launch Claude Code, Cursor, or your preferred AI coding agent.

Try It Out

Use the prompt template or examples below to test the skill.

Customize

Adapt the skill to your specific use case and workflow.

Usage Examples

Prompt Template

Design a C3 architecture for a [PRODUCT_TYPE] application. The application should have [NUMBER] core features: [FEATURE_LIST]. Generate the context diagram, container diagram, and component diagrams with detailed descriptions. Use [PROGRAMMING_LANGUAGE] for the implementation.

Example Output

For a task management application with 5 core features: task creation, task assignment, task tracking, reporting, and user management, here's a C3 architecture design:

1. Context Diagram:
   - System: Task Management Application
   - Users: Project Managers, Team Members, Admins
   - Other Systems: Email Service, Calendar Service, Database

2. Container Diagram:
   - Web Container: Handles user interfaces and business logic (React.js)
   - API Container: Manages data processing and business logic (Node.js)
   - Database Container: Stores application data (PostgreSQL)

3. Component Diagrams:
   - Task Creation Component: Handles task creation and validation
   - Task Assignment Component: Manages task assignment and notifications
   - Task Tracking Component: Tracks task progress and updates
   - Reporting Component: Generates reports and analytics
   - User Management Component: Manages user authentication and authorization

Each component is designed to be modular and reusable, following the C3 architecture methodology. The application is implemented in JavaScript, with React.js for the frontend and Node.js for the backend.