academic-paper-composer

Writes academic papers from detailed outlines with chapter-level and final quality evaluation.

11|2|Updated May 17, 2026
One-click install
npx skills add https://github.com/haozhang04/homework_skill --skill academic-paper-composer-haozhang04
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: academic-paper-composer
Source: https://github.com/haozhang04/homework_skill/tree/main/skill/academic_paper_skills/composer
Command: npx skills add https://github.com/haozhang04/homework_skill --skill academic-paper-composer-haozhang04

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? Turning a detailed paper outline into a complete, submission-ready manuscript is slow and error-prone: chapters drift in style, citations get missed, and quality issues surface only at the end. This Skill executes an outline chapter-by-chapter with built-in quality gates so problems are caught and fixed at each stage. ## Core Features & Use Cases - Systematic Chapter Writing: Writes each chapter following the outline's structure, word counts, and citation plan, with a 5-dimension quality check (scored out of 20, pass threshold 16) after every chapter. - Final 7-Dimension Evaluation: Runs a comprehensive 70-point assessment covering argument quality, literature integration, originality, methodology, structure, and platform conformity, plus a completeness checklist before submission. - Submission Package Preparation: Produces platform-specific checklists for PhilArchive, PhilPapers, arXiv, and PhilSci-Archive along with the final manuscript, abstract, and metadata. - Use Case: After using academic-paper-strategist to produce an optimized outline for a philosophy paper, use this Skill to write all chapters, validate each one, and receive a submission-ready manuscript with quality reports. ## Quick Start Ask the AI to write the full paper from your optimized outline, evaluating each chapter with the quality check scripts before moving to the next.

Frequently Asked Questions about academic-paper-composer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I write an academic paper from an outline?

Provide a detailed outline with chapter titles, word counts, content guidance, and key citations. The Skill writes each chapter in sequence, runs a 5-dimension quality check after each one, and only proceeds when the chapter scores at least 16 out of 20.

What input does academic-paper-composer require?

It requires an optimized detailed outline specifying chapter structure to the third level, word counts, content guidance, and key citations. A platform writing standards guide and a literature base are recommended. The academic-paper-strategist skill produces this outline, but any equivalent outline works.

How does the chapter quality check script work?

Run chapter_quality_check.py to create a JSON evaluation template, fill in scores for five dimensions, then generate a report. It calculates the total out of 20, flags dimensions scoring below 3, checks word count variance, and outputs prioritized revision recommendations.

What happens if a chapter fails the quality gate?

Chapters scoring below 16 out of 20 must be revised before proceeding. The report identifies weak dimensions and suggests targeted fixes, such as adding citations, clarifying prose, or restructuring. You re-evaluate after revisions until the chapter passes.

Which submission platforms does the final evaluation support?

The final evaluation script includes platform-specific checklists for PhilArchive, PhilPapers, arXiv, and PhilSci-Archive, covering format, abstract length, metadata, and classification requirements. Other platforms get a generic verification checklist.

What are the limitations of systematic outline-based writing?

Quality depends heavily on outline detail; vague outlines produce weaker manuscripts. The process takes one to two days for a typical paper due to iterative checks, and final human review is still recommended since automated scoring cannot catch every issue.