SuperAPI AI Directory

lovstudio-thesis-polish

Testing & quality

View repository

Updated 10-16 00:01

About this skill

thesis-polish — Comprehensive polishing of MBA thesis

Elevate an MBA thesis from draft to national outstanding thesis standard. Covers four dimensions: language polishing, structural optimization, argument strengthening, and highlighting innovation points.

When to Use

  • The user provides an MBA thesis (in markdown format) that needs polishing
  • The user wants the thesis to reach the national outstanding thesis standard
  • The user needs comprehensive improvement of academic writing

Workflow (MANDATORY)

Step 1: Receive the thesis

Read the thesis file provided by the user (in markdown format). If the user provides a file path, use the Read tool to read it.

Step 2: Diagnostic assessment

After reading the full text, first output adiagnostic report(no more than 300 words), including:

Dimension Current level Main problems Direction for improvement
Language expression A/B/C ... ...
Argumentation logic A/B/C ... ...
Structural layout A/B/C ... ...
Innovation contribution A/B/C ... ...
Use of literature A/B/C ... ...

Step 3: Confirm the polishing strategy

Use AskUserQuestion to confirm strategy BEFORE polishing.

After presenting the diagnostic report to the user, ask:

  • Do you agree with the diagnosis? Are there any special requirements?
  • What are the core innovation points of the thesis? (To help the AI highlight them better)
  • Are there any special requirements from the target journal/review standards?

Step 4: Polish chapter by chapter

Polish chapter by chapter, outputting the complete polished text for each chapter. Strictly follow the standards below when polishing:

4.1 Language expression standards

  • Academic terminology: colloquial expressions → standard academic terminology ("a lot" → "a large number of/numerous", "feel" → "believe/indicate")
  • Sentence upgrading: simple sentences → compound sentences, adding logical connections of cause, contrast, and progression
  • Terminology consistency: unify terminology throughout the text, and mark the original English term at first occurrence
  • Quantified expression: vague descriptions → precise data/proportions/citation support
  • Passive and active: use more passive voice in the research methods section, and active discussion in the discussion section

4.2 Argumentation logic standards

  • Claim-evidence-argumentation: every viewpoint must be supported by data/literature/cases
  • Logical chain: there must be a clear logical progression between preceding and following paragraphs
  • Anticipating rebuttals: respond in advance to possible doubts ("although... but...")
  • Comparative argumentation: compare with existing research to highlight the contribution of this paper
  • Causal rigor: distinguish correlation from causality, and avoid over-inference

4.3 Structural layout standards

  • Chapter transitions: each chapter begins with a transitional paragraph that connects the previous and following content
  • Paragraph structure: topic sentence → elaboration → brief summary, with each paragraph focusing on one point
  • Clear hierarchy: first-level heading (chapter) → second-level heading (section) → third-level heading (subsection), with no more than three levels
  • Abstract improvement: the abstract covers the five elements of research background, purpose, methods, findings, and contribution

4.4 Innovation Contribution Standards

  • Explicit labeling: Clearly list 2-3 innovation points in the introduction and conclusion
  • Differentiated expression: Use the sentence pattern "Different from existing research... this paper is the first to/innovatively..."
  • Theoretical contribution: Explain the expansion or revision of the theoretical framework
  • Practical significance: Explain the specific guiding value for management practice

4.5 Literature Use Standards

  • Authoritativeness: Prioritize citing SSCI/CSSCI journals and classic works
  • Timeliness: The proportion of literature from the past 5 years should be no less than 60%
  • Sense of dialogue: Do not just list the literature, but "dialogue" with it ("XXX (2023) pointed out..., and on this basis this paper further finds...")
  • Literature review: Organize by theme/school rather than piling up by time

Step 5: Output

Merge all polished chapters and output the complete polished full text (in markdown format).

If the user needs to save it as a file, use the Write tool to write it, with the file name following: 手工川-mba-thesis-polished-{YYYY-MM-DD}-v0.1.md

Important Notes

  • Do not tamper with data: Polishing only involves expression and argumentation methods; do not fabricate data or invent citations
  • Preserve the original meaning: All modifications must be faithful to the author's original meaning and must not change the research conclusions
  • Annotation suggestions: If an argument somewhere is weak and the author needs to supplement data/literature, use [建议补充: ...] to annotate
  • Process in batches: Papers are usually very long; polish in batches by chapter to avoid omissions