Paper CT Seven-Dimension Flaw Scanning System

Problem

Before submission, papers often have hidden weaknesses: inconsistent formatting, unverifiable citations, mismatched data, weak feasibility, or obvious AI-style writing. Manual review can miss categories, while grammar tools do not cover academic credibility or journal fit. Paper CT Seven-Dimension Flaw Scanning System is aimed at Chinese academic papers and turns review into a measurable, local, report-driven process.

How It Works

It follows the CT three-step method: penetrate the surface text, diagnose concrete flaws, and issue prescriptive revision guidance. It reads the paper locally and evaluates seven dimensions:
- Formatting: fonts, headings, line spacing, figure/table numbering, and GB/T 7714 reference style.
- Factual integrity: data sources, citation existence, and internal consistency.
- Feasibility: method validity, sample size, timeline, and resource fit.
- Authenticity: cross-validation, statistical plausibility, and logical fallacies.
- AI traces: uniform paragraph rhythm, vague language, formulaic transitions, and suspicious references.
- Academic norms: plagiarism signals, ethics, reference completeness, and funding disclosure.
- Journal fit: scope, length, level, and structural completeness.

It outputs human-readable report_summary.md, machine-readable report_detail.json, and a score overview in scores.json. Errors are not silently ignored: read failures, missing dependencies, and module failures are recorded with actionable context.

Boundaries

It is useful for pre-submission self-checks, locating revision priorities, and producing reviewable score reports. It does not replace plagiarism checking, grammar correction, English polishing, ghostwriting, or guarantee acceptance. It is designed for academic texts, not novels, official documents, or contracts.