writerkit-init

name	/writerkit-init
description	Use when the user wants to start, plan, write, or publish technical and scientific books in LaTeX, requiring end-to-end orchestration of the production lifecycle.
metadata	{"triggers":"technical books, LaTeX, scientific writing, book production, academic references"}

Producing Technical Books (WriterKit)

Overview

This skill orchestrates the full lifecycle of technical book production, from initial setup and planning to final publication. It manages a sequence of specialized sub-skills to ensure academic rigor, technical accuracy, and LaTeX compatibility.

When to Use This Skill

• Starting a new technical or scientific book project.
• Producing a complete chapter following a structured workflow.
• Ensuring code examples are correctly rendered in LaTeX (accent mapping).
• Validating academic references against real-world databases.

How to Use It

The workflow is divided into two main stages: Book Initialization & Planning, and Chapter Production.

Stage I: Book Initialization & Planning

0. Initialization (/writerkit-init): MANDATORY FIRST STEP.
   • Copy the ~/.gemini/antigravity/skills/writerkit/writerkit-init/memory/ directory to the local project root as ./memory/.
   • Create the following project structure:
     • ./chapters: For chapter .tex files.
     • ./codes: For code snippet .tex files, organized by ./codes/chapterxx/.
     • ./images: For png, svg, and tikz files, organized by ./images/chapterxx/.
     • ./fonts: For custom font files (defaults to LaTeX fonts if empty).
     • ./configs: For .cls and .sty configuration files.
     • ./templates: For user-provided template models.
     • ./output: To hold the final compilable structure for TexPage.
   • Initialize a main.tex in the root with capa, TOC, and basic imports.
1. Setup & Book Planning (/writerkit-setup):
   • Initialize project and editorial pillars (tone, language, depth) in ./memory/memory.json.
   • Run LaTeX template DNA extraction.
   • Plan the entire book outline (Table of Contents / TOC and list of chapters) and save it in the project memory.
2. User Approval Gate (MANDATORY):
   • Present the overall book plan and planned TOC to the user.
   • Stop and request explicit approval from the user. Do NOT proceed to individual chapter production until approval is granted and logged.

Stage II: Chapter Production Workflow

3. Chapter Orchestration (/writerkit-chapter):
   • For each chapter in the approved plan, execute the rigid workflow defined in workflow/orchestrator.md:
     1. Planning (/writerkit-planning): Define chapter objectives, outline sections, identify code examples.
     2. Drafting (/writerkit-drafting): Use chapter template, write conceptual content, add placeholders.
     3. Code Development (/writerkit-code): Implement and test code examples externally (mapping accents).
     4. Integration (/writerkit-integration): Insert validated code into chapter, add explanations.
     5. Review (/writerkit-review): Self-review using the production checklist.
     6. Validation (/writerkit-validation): Run validation scripts (including the Reference Checker).
     7. Revision (/writerkit-revision): Address issues and improve clarity.
     8. Publication (/writerkit-publication): Finalize and publish chapter.

Mandatory Rules

• Structural Integrity: ALWAYS follow the defined folder schema. Never place assets outside their designated directories.
• Progressive Loading: Activate the specific sub-skill for each phase.
• Local Project Memory: All state and decisions MUST be recorded in the local ./memory/memory.json file. NEVER write to the global skill path.
• Decision Logging: Every transition and decision must be recorded in ./memory/memory.json.
• Main File Logic: main.tex must be the orchestrator of the book, importing styles from ./configs and chapters from ./chapters.
• Accent Mapping: Always ensure literate mapping for non-ASCII characters in code blocks.
• Reference Integrity: Use the Reference Checker script during the Validation phase.