create-agentic-lemon 0.1.0 → 0.1.1

package/README.md

@@ -1,61 +1,140 @@
-# create-agentic-lemon
-
-An agent-agnostic workspace generator that creates a reusable AI operating system folder structure.
-
-Treat your AI assistant context as files and folders instead of transient, ephemeral chat history. This workspace structure helps any file-reading AI coding tool (ChatGPT, Claude, Cursor, Windsurf, Copilot, etc.) understand your project context, guidelines, workflows, decisions, and learnings instantly.
-
-## Usage
-
-You can generate a workspace with a single command without installing the package:
-
+# 🍋 create-agentic-lemon
+
+An agent-agnostic workspace generator that scaffolds a standardized, persistent **AI Operating System (OS)** folder structure within your codebase.
+
+---
+
+## Why Agentic Lemon?
+
+AI assistants (ChatGPT, Claude, Cursor, Windsurf, Copilot, Cline, Antigravity, etc.) are incredibly powerful but limited by transient chat histories and context windows. Every new chat session requires you to re-explain your system architecture, project constraints, database schemas, and coding preferences. Worse, when they make a mistake and you correct it, they forget it in the next session.
+
+**Agentic Lemon** solves this by treating your AI assistant instructions, project context, playbooks, memory, and references as structured files in your project directory. This enables any file-reading AI coding tool to align with your project context, guidelines, workflows, and past learnings instantly.
+
+---
+
+## Core Features
+
+*   🛠️ **Agent-Agnostic Design:** Works seamlessly out-of-the-box with any modern IDE, chat interface, or agentic framework that can read workspace files.
+*   🧠 **Persistent Memory & Learnings:** Dedicated structures to store architectural decisions, developer preferences, and recurring corrections, preventing the AI from repeating the same mistake twice.
+*   📋 **Repeatable Workflows (Skills):** Standardized workflow playbooks (for software development, debugging, database queries, documentation, and reports) complete with checklists, examples, templates, and learnings.
+*   📁 **Clean Output Sandbox:** Structured directories for AI-generated drafts, approved snippets, and debug logs to keep your main codebase clean.
+
+---
+
+## Quick Start
+
+Generate your workspace in seconds with a single command without installing the package:
+
 ```bash
 npm create agentic-lemon@latest my-workspace
 ```
-
-Or using `npx`:
-
+
+Or using `npx`:
+
 ```bash
 npx create-agentic-lemon@latest my-workspace
 ```
 
-Or install it globally and use either command alias:
+### Installation as a Global CLI
+
+You can also install the tool globally:
 
 ```bash
 npm i -g create-agentic-lemon
+```
+
+Then run it using either command alias:
+
+```bash
 agentic-lemon my-workspace
+# OR
 create-agentic-lemon my-workspace
 ```
-
-### Options
-
-*   `--force`: Overwrite existing template files in the workspace (leaves other files intact).
-
----
-
-## Workspace Structure
-
-The generated workspace structure includes:
-
-*   **`AGENTS.md`**: Global operating rules for the AI.
-*   **`PROJECT_CONTEXT.md`**: High-level metadata about project scope, users, systems, constraints, and styling preferences.
-*   **`context/`**: Stable knowledge files (e.g. database schema, systems documentation, coding styles).
-*   **`skills/`**: Repeatable workflows documented as reusable playbooks (e.g. software development, debugging, reports).
-*   **`memory/`**: High-level cross-skill memory files including decisions, preferences, and recurring corrections.
-*   **`outputs/`**: Dedicated subdirectories for drafts, approved content, snippets, and debug notes.
-*   **`references/`**: Links to external documentation and sources.
-
----
-
-## How to use with an AI Assistant
-
-Open the generated workspace in your editor. In your AI chat window, prompt the AI:
-
-> "Read AGENTS.md first, then PROJECT_CONTEXT.md, then any relevant files under context/ and skills/ before answering."
-
-When the AI makes a mistake or you correct something reusable, save that learning in the respective skill's `learnings.md` or global `memory/recurring-corrections.md` so the AI remembers it next time!
-
----
-
-## License
-
-MIT
+
+### Options
+
+*   `--force`: Overwrite existing template files in the workspace (leaves other custom files intact).
+*   `-h, --help`: Show help and usage instructions.
+
+---
+
+## Workspace Architecture
+
+Here is the standardized folder structure generated by **Agentic Lemon**:
+
+```mermaid
+graph TD
+    Root[Workspace Root] --> AGENTS["AGENTS.md (Global Rules)"]
+    Root --> PROJ["PROJECT_CONTEXT.md (Project Specs)"]
+    
+    Root --> Context["context/ (Stable Knowledge)"]
+    Context --> Arch["architecture.md"]
+    Context --> Coding["coding-style.md"]
+    Context --> Const["constraints.md"]
+    Context --> DB["database-schema.md"]
+    Context --> Gloss["glossary.md"]
+    Context --> Overview["project-overview.md"]
+    Context --> Sys["systems.md"]
+    
+    Root --> Skills["skills/ (Workflows & Playbooks)"]
+    Skills --> SoftDev["software-development/"]
+    SoftDev --> Skill["skill.md"]
+    SoftDev --> Checklist["checklist.md"]
+    SoftDev --> Examples["examples.md"]
+    SoftDev --> Learnings["learnings.md"]
+    SoftDev --> Templates["templates.md"]
+    Skills --> DBWork["database-work/"]
+    Skills --> Debugging["debugging/"]
+    Skills --> Doc["documentation/"]
+    Skills --> Report["report-generation/"]
+    
+    Root --> Memory["memory/ (Persistent Memory)"]
+    Memory --> Changelog["changelog.md"]
+    Memory --> Decisions["decisions.md"]
+    Memory --> Prefs["preferences.md"]
+    Memory --> Corrs["recurring-corrections.md"]
+    
+    Root --> Outputs["outputs/ (Deliverables & Temp Files)"]
+    Outputs --> Drafts["drafts/"]
+    Outputs --> Appr["approved/"]
+    Outputs --> Snip["snippets/"]
+    Outputs --> Debug["debug-notes/"]
+    
+    Root --> References["references/ (External Material)"]
+    References --> ExtDocs["external-docs.md"]
+    References --> Links["links.md"]
+    References --> Sources["sources.md"]
+```
+
+### Directory Structure Details
+
+*   **`AGENTS.md`**: Global operating rules, behavior guidelines, and constraints for the AI.
+*   **`PROJECT_CONTEXT.md`**: High-level metadata about project scope, users, tech stack, and styling preferences.
+*   **`context/`**: Slow-moving, stable knowledge files (e.g., database schema, systems architecture, coding styles).
+*   **`skills/`**: Repeatable workflows documented as reusable playbooks. Each skill includes custom checklists, templates, examples, and learnings.
+*   **`memory/`**: High-level cross-skill memory files including design decisions, project preferences, and recurring corrections.
+*   **`outputs/`**: Dedicated subdirectories for drafts, approved content, snippets, and debug logs.
+*   **`references/`**: Links to external documentation, specifications, and third-party APIs.
+
+---
+
+## How to use with an AI Assistant
+
+### 1. Bootstrap
+Generate the workspace structure at the root of your project:
+```bash
+npx create-agentic-lemon@latest .
+```
+
+### 2. System Boot Prompt
+Whenever you start a new conversation with your AI assistant, instruct it to load the operating system files:
+> "Read AGENTS.md first, then PROJECT_CONTEXT.md, then any relevant files under context/ and skills/ before answering."
+
+### 3. The Learning Loop
+When you correct the AI's behavior, or establish a new architectural pattern, save that learning in the respective skill's `learnings.md` or the global `memory/recurring-corrections.md` file. The AI will read it at the start of future sessions, preventing repeated mistakes!
+
+---
+
+## License
+
+This project is licensed under the MIT License.

package/package.json

@@ -1,7 +1,7 @@
-{
-  "name": "create-agentic-lemon",
-  "version": "0.1.0",
-  "description": "Create an agent-agnostic AI operating system folder structure.",
+{
+  "name": "create-agentic-lemon",
+  "version": "0.1.1",
+  "description": "Create an agent-agnostic AI operating system folder structure.",
   "type": "module",
   "bin": {
     "agentic-lemon": "bin/agentic-lemon.js",