Can I use this for technical documentation?

Absolutely. It is ideal for large-scale documentation projects where you need to ensure terminology, tone, and structure are consistent across dozens of separate files.

Does this require a separate API key?

Yes, you need a valid GEMINI_API_KEY set in your environment variables to utilize the Gemini 2.5 Pro model capabilities through this script.

How does it handle author annotations?

You can explicitly instruct the skill via the prompt file to interpret specific symbols (like curly braces) as editorial directives to be executed rather than text to be rewritten.

Why use Gemini Writer instead of Claude directly?

While Claude is highly capable, Gemini's 1M token window allows the AI to ingest an entire book manuscript and all editorial notes at once, ensuring global consistency that isn't possible with smaller context windows.

What is the benefit of Batch Mode?

Batch mode instructs Gemini to look at the entire manuscript but only output one chapter at a time. This prevents output truncation and ensures the highest quality writing for each section while maintaining full context.

Gemini Writer

Name: Gemini Writer
Author: cdeistopened

bycdeistopened

•

Gestión de Contenido

Delegates large-context writing and manuscript-scale refactoring to Gemini's 1-million token window.

Gemini Writer is a specialized tool for Claude Code designed to handle massive writing projects that exceed typical context limits. By leveraging Gemini 2.5 Pro's 1-million token window, this skill allows users to process entire book manuscripts, perform structural refactoring, and conduct redundancy audits across an entire body of work in a single pass. Its unique batch processing mode enables the AI to maintain full-manuscript awareness while generating high-quality, chapter-by-chapter outputs, ensuring consistent voice and thematic coherence throughout the entire document.

Características Principales

01Advanced voice-matching through system-level style instructions

028 GitHub stars

031M token context window for processing entire manuscripts at once

04Batch processing mode for high-quality, chapter-by-chapter generation

05Structural refactoring to resolve redundancies and reorder content

06Dry-run capability to verify token counts and costs before execution

Casos de Uso

01Performing a comprehensive redundancy analysis across a large documentation set

02Applying global editorial annotations and voice guides to multiple files simultaneously

03Rewriting a full-length book manuscript while maintaining cross-chapter consistency

Características Principales

01Advanced voice-matching through system-level style instructions

028 GitHub stars

031M token context window for processing entire manuscripts at once

04Batch processing mode for high-quality, chapter-by-chapter generation

05Structural refactoring to resolve redundancies and reorder content

06Dry-run capability to verify token counts and costs before execution

Casos de Uso

01Performing a comprehensive redundancy analysis across a large documentation set

02Applying global editorial annotations and voice guides to multiple files simultaneously

03Rewriting a full-length book manuscript while maintaining cross-chapter consistency