A multi-agent writing system designed to iteratively process and refine text through specialized editing agents. Each agent focuses on a specific aspect of writing improvement, working together to produce polished, professional content.
This system processes text through a series of specialized agents, each with a specific role in improving different aspects of writing quality. The agents work in sequence, with each agent's output becoming the input for the next agent in the pipeline.
Purpose: Surface-level corrections
Function: Corrects typos, spelling mistakes, punctuation, capitalization, and spacing while preserving the author's voice and meaning.
Purpose: Regional language standardization
Function: Converts text to UK English standards, including spelling, vocabulary, punctuation, and formatting while maintaining meaning and tone.
Purpose: Structural flow and readability
Function: Improves logical progression, readability, and consistency of tone while ensuring smooth transitions between sections.
4. Headings
Purpose: Document structure
Function: Generates, reviews, and refines subheadings to create clear, logical sections that support readability and comprehension.
5. Proofreader
Purpose: Fact verification and source linking
Function: Verifies facts and statistics, corrects inaccuracies, and inserts precise inline Markdown links for claims requiring sources.
6. Add Sources
Purpose: Source integration
Function: Integrates reliable sources by inserting inline Markdown links at factual claims while preserving document structure and tone.
7. SEO Reviewer
Purpose: Search engine optimization
Function: Applies minimal, high-precision edits to improve organic search performance without altering flow, voice, or narrative structure.
The recommended processing order is:
- Basic Typo Reviewer - Clean up surface-level errors
- UK English Standardiser - Standardize to UK English
- Flow and Polish - Improve readability and logical flow
- Headings - Structure the document with clear headings
- Proofreader - Verify facts and add sources
- Add Sources - Enhance source integration
- SEO Reviewer - Final SEO optimization
Each agent configuration can be used individually by copying the prompt from the respective markdown file and applying it to your text.
Use the process_text.py script to automatically process text through the entire agent pipeline:
python3 process_text.pyThis script will:
- Read text from
input.md - Process it through each agent in sequence
- Save the final result to
output.md - Save intermediate results for review
Claude-Code-Writing-Squad/
├── README.md # This file
├── agents/ # Agent configuration files
│ ├── basic-typo-reviewer.md
│ ├── uk-english-standardiser.md
│ ├── flow-and-polish.md
│ ├── headings.md
│ ├── proofreader.md
│ ├── add-sources.md
│ └── seo-reviewer.md
├── process_text.py # Automated processing script
├── input.md # Input text file
└── output.md # Final processed output
- Autonomous Operation: All agents work independently without requiring user approval
- Meaning Preservation: Each agent maintains the original intent and message
- Collaborative Design: Agents are designed to work together in sequence
- Markdown Focus: All agents handle Markdown formatting appropriately
- Direct Output: Agents return processed text without commentary or explanations
- Place your text in
input.md - Run
python3 process_text.py - Review the processed output in
output.md - Check intermediate files if you need to see the progression through each agent
When adding new agents or modifying existing ones:
- Ensure agents work autonomously without user interaction
- Maintain compatibility with Markdown formatting
- Preserve meaning and authorial intent
- Follow the established pattern of direct text output
- Update this README with any new agents or changes to the processing flow