Problem it solves
Many academic-writing prompts optimize only the final prose. This suite treats a paper as a research pipeline in which claims depend on sources, methods, analyses, and reporting choices. Agents must distinguish full-text verification, abstract or metadata verification, user-provided evidence, analysis output, inference, and unverified material.
What is included
- 1 orchestrator skill + 12 specialist skills, with
write-academic-paperas the multi-stage entry point; - Covers research-question definition, literature review, study design, data analysis, thesis and journal-article writing, defense, submission, peer-review revision, and research-integrity audit;
- Built for ChatGPT and Codex, following the open Agent Skills folder format (each skill is a self-contained directory anchored by a
SKILL.md).
Scope
The suite does not fabricate citations, DOIs, page numbers, data, or reviewer comments. It maintains claim–evidence alignment and weakens or flags unsupported claims. Discipline-specific reporting standards, terminology, and methods must be supplied as project context. It does not replace domain expertise, ethical approval, statistical review, source verification, or authorship accountability.
Repository
→ https://github.com/BettyXu2001/awesome-academic-writing-skills