@lylll9436/paper-polish-workflow-skill 2.1.0

package/examples/abstract-polishing-session.md

@@ -0,0 +1,263 @@
+# Example Session: Polishing an Abstract
+
+This document shows a complete session of polishing an abstract for an academic journal submission.
+
+## Context
+
+- **Paper**: Deep learning framework for medical image classification (Fictional Example)
+- **Target Journal**: Example Journal (with ≤8,000 words, ≤250 words abstract)
+- **Section**: Abstract
+
+> **Note**: This is a fictional example created for demonstration purposes only. All research details, methods, and results are fabricated.
+
+---
+
+## Phase 0: Pre-Polish Baseline
+
+**AI**: 
+```
+Before we start polishing, I need some information:
+1. What is the target journal? Any special requirements?
+2. Where are the example/reference papers?
+3. Which section should we start with?
+```
+
+**User**: Target journal has 250-word abstract limit, examples in docs/example, start with abstract
+
+**AI**: 
+```
+## Pre-Polish Checklist
+- Target: Example Journal (≤8,000 words total, ≤250 words abstract)
+- Example papers: docs/example/*.pdf
+- Section: Abstract
+- Current abstract word count: ~280 words (needs trimming)
+```
+
+---
+
+## Step 1: Confirm Macro Structure
+
+**AI**:
+```markdown
+## Abstract Structure Analysis
+
+| Component | Function | Suggested Sentences |
+|-----------|----------|---------------------|
+| Background | Research importance | 1-2 |
+| Gap | Existing limitations | 1-2 |
+| Method | Proposed approach | 3-4 |
+| Results | Key findings | 1-2 |
+| Contribution | Significance | 1 |
+
+Please confirm this structure, or describe adjustments needed.
+```
+
+**User**: The structure looks good. Confirm.
+
+---
+
+## Step 2: Confirm Per-Sentence Logic
+
+**AI**:
+```markdown
+## Per-Sentence Logic Mapping
+
+| Sentence | Logic Function | Key Points |
+|----------|----------------|------------|
+| S1 | Background | Medical imaging importance for clinical diagnosis |
+| S2 | Data | Large-scale datasets enabling AI-based analysis |
+| S3 | Gap | Current methods → limited interpretability → hard to trust → clinical adoption barrier |
+| S4 | Method Overview | Hybrid framework: local features + global context |
+| S5 | Method Detail | CNN extracts local patterns from image regions |
+| S6 | Method Detail | Attention mechanism highlights diagnostic regions |
+| S7 | Method Detail | Multi-scale fusion combines different resolution features |
+| S8 | Method Detail | Interpretability module generates visual explanations |
+| S9 | Results | Evaluated on benchmark datasets, achieves competitive accuracy |
+| S10 | Contribution | Accuracy + interpretability + clinical applicability |
+
+Please confirm each sentence's logic.
+```
+
+**User**: 
+- S1: Confirm
+- S2: Confirm  
+- S3: Modify - emphasize that interpretability is essential for clinical trust
+- S4-S8: Confirm
+- S9: Confirm
+- S10: Confirm
+
+**AI**: Updated S3 logic to: "Current methods → limited interpretability (essential for clinical trust) → adoption barrier in real-world settings"
+
+---
+
+## Step 3: Confirm Specific Expressions
+
+**AI** uses `mcp_question`:
+
+```
+┌─────────────────────────────────────────────────────────────┐
+│ S1 Expression                                               │
+│ ─────────────────────────────────────────────────────────── │
+│ S1: Medical imaging importance for clinical diagnosis       │
+│                                                             │
+│ ○ A: Medical imaging plays a critical role in modern        │
+│      clinical diagnosis, disease monitoring, and treatment  │
+│      planning.                                              │
+│                                                             │
+│ ○ B: Advances in medical imaging have revolutionized        │
+│      clinical practice, enabling non-invasive diagnosis     │
+│      across specialties.                                    │
+│                                                             │
+│ ○ C: Accurate interpretation of medical images is           │
+│      fundamental to clinical decision-making and patient    │
+│      outcomes.                                              │
+│                                                             │
+│ ○ Type your own answer                                      │
+└─────────────────────────────────────────────────────────────┘
+```
+
+**User selects**: A
+
+```
+┌─────────────────────────────────────────────────────────────┐
+│ S2 Expression                                               │
+│ ─────────────────────────────────────────────────────────── │
+│ S2: Large-scale datasets enabling AI analysis               │
+│                                                             │
+│ ○ A: The availability of large-scale annotated datasets     │
+│      has accelerated the development of AI-based diagnostic │
+│      systems.                                               │
+│                                                             │
+│ ○ B: Deep learning methods have shown remarkable success    │
+│      in automated image analysis, driven by growing data    │
+│      availability.                                          │
+│                                                             │
+│ ○ C: Recent advances in deep learning, combined with        │
+│      large medical imaging datasets, have enabled automated │
+│      diagnostic tools.                                      │
+│                                                             │
+│ ○ Type your own answer                                      │
+└─────────────────────────────────────────────────────────────┘
+```
+
+**User selects**: C
+
+*(Process continues for all sentences...)*
+
+---
+
+## Step 4: Reference Paper Consultation
+
+**User**: "Is 'clinical decision-making and patient outcomes' the right terminology?"
+
+**AI** analyzes example papers:
+
+```markdown
+## Reference Paper Patterns
+
+| Paper | Opening Phrase | Pattern |
+|-------|---------------|---------|
+| Paper 1 | "diagnosis and treatment planning" | task-focused |
+| Paper 2 | "clinical workflow and patient care" | process-focused |
+| Paper 3 | "diagnostic accuracy and efficiency" | outcome-focused |
+| Paper 4 | "clinical practice and decision support" | application-focused |
+
+## Suggested Alternatives
+
+| Option | Expression |
+|--------|------------|
+| A | clinical decision-making and patient outcomes |
+| B | diagnostic accuracy and clinical efficiency |
+| C | clinical workflow and patient care |
+| D | clinical decision-making and patient outcomes (keep original) |
+```
+
+**User selects**: D (keep original)
+
+---
+
+## Step 5: Repetition & Coherence Check
+
+**AI**:
+```markdown
+## Repetition Check
+
+| Location | Issue | Suggestion |
+|----------|-------|------------|
+| S4 vs S10 | Both mention "interpretability" | Rephrase S4 to "explainable predictions" |
+
+## Coherence Check
+
+| Location | Issue | Suggestion |
+|----------|-------|------------|
+| S3→S4 | Abrupt transition from gap to method | Add transition word |
+```
+
+**AI** uses `mcp_question` for transition:
+
+```
+┌─────────────────────────────────────────────────────────────┐
+│ Transition Word                                             │
+│ ─────────────────────────────────────────────────────────── │
+│ S3→S4 transition: From 'problem' to 'solution'              │
+│                                                             │
+│ ○ A: "To address these limitations, this study..."          │
+│ ○ B: "In response, this study..."                           │
+│ ○ C: "To bridge this gap, we..."                            │
+│ ○ D: Keep as is (no transition)                             │
+└─────────────────────────────────────────────────────────────┘
+```
+
+**User selects**: A
+
+---
+
+## Final Version
+
+**AI**:
+```markdown
+## Final Abstract
+
+Medical imaging plays a critical role in modern clinical diagnosis, 
+disease monitoring, and treatment planning. Recent advances in deep 
+learning, combined with large medical imaging datasets, have enabled 
+automated diagnostic tools. However, existing methods often lack 
+interpretability, which is essential for clinical trust and limits 
+their adoption in real-world settings. To address these limitations, 
+this study introduces a hybrid framework combining local feature 
+extraction with global context modeling. The framework employs 
+convolutional neural networks to extract local patterns from image 
+regions, while an attention mechanism highlights diagnostically 
+relevant areas. Multi-scale feature fusion integrates information 
+across resolutions, and an interpretability module generates visual 
+explanations for predictions. We evaluate the framework on multiple 
+benchmark datasets, achieving competitive accuracy while providing 
+clinician-friendly explanations. The framework advances both 
+diagnostic accuracy and interpretability, supporting practical 
+clinical deployment.
+
+**Word Count**: ~150 words ✅ (within 250 limit)
+
+Confirm to write to file?
+```
+
+**User**: Yes
+
+**AI**: ✅ Written to `docs/paper/sections/00_abstract_polished.md`
+
+---
+
+## Summary
+
+| Step | Action | User Input |
+|------|--------|------------|
+| Phase 0 | Collect requirements | Journal specs, docs/example |
+| Step 1 | Confirm structure | Confirmed |
+| Step 2 | Confirm sentence logic | Modified S3 |
+| Step 3 | Select expressions | A, C, B, B, ... |
+| Step 4 | Reference consultation | Keep original |
+| Step 5 | Coherence check | Added transition |
+| Write | Final confirmation | Yes |
+
+**Total conversation turns**: ~15
+**Time saved**: Systematic approach avoids back-and-forth revisions

package/package.json

@@ -0,0 +1,38 @@
+{
+  "name": "@lylll9436/paper-polish-workflow-skill",
+  "version": "2.1.0",
+  "description": "A 14-Skill suite (ppw:* namespace) for academic paper writing, repo-to-paper generation, and submission — powered by Claude Code.",
+  "keywords": [
+    "claude-code",
+    "claude-code-skills",
+    "academic-writing",
+    "paper-polishing",
+    "latex",
+    "ai-assisted-writing"
+  ],
+  "author": "Lylll9436",
+  "license": "MIT",
+  "repository": {
+    "type": "git",
+    "url": "git+https://github.com/Lylll9436/Paper-Polish-Workflow-skill.git"
+  },
+  "homepage": "https://github.com/Lylll9436/Paper-Polish-Workflow-skill#readme",
+  "bugs": {
+    "url": "https://github.com/Lylll9436/Paper-Polish-Workflow-skill/issues"
+  },
+  "publishConfig": {
+    "registry": "https://registry.npmjs.org",
+    "access": "public"
+  },
+  "files": [
+    "paper-polish-workflow/",
+    ".claude/skills/",
+    "references/",
+    "examples/",
+    "CHANGELOG.md",
+    "CONTRIBUTING.md",
+    "CONTRIBUTING_CN.md",
+    "LICENSE",
+    "README.md"
+  ]
+}

package/paper-polish-workflow/SKILL.md

@@ -0,0 +1,160 @@
+---
+name: paper-polish-workflow
+description: Systematic top-down workflow for polishing academic papers. Structure to logic to expression with user confirmation at each step.
+---
+
+## Purpose
+
+This Skill provides a systematic, top-down workflow for polishing academic papers. It works from structure to logic to expression, with user confirmation at each decision point. Expression options are drawn from reference-driven academic patterns rather than ad hoc rewrites, ensuring professional and consistent output suitable for journal submission.
+
+## Trigger
+
+**Activates when the user asks to:**
+- Polish, revise, or improve an academic paper section by section
+- 润色、精修、逐步改进学术论文
+
+**Example invocations:**
+- "Polish my paper section by section" / "润色论文"
+- "Help me revise my introduction step by step" / "精修论文"
+- "Guide me through polishing this draft" / "帮我逐步润色这篇论文"
+
+## Modes
+
+| Mode | Default | Behavior |
+|------|---------|----------|
+| `interactive` | Yes | Full 4-step flow with user confirmation at each decision point |
+| `guided` | | Multi-pass with confirmation at key checkpoints only |
+| `direct` | | Single-pass polish using defaults; skip AskUserQuestion |
+| `batch` | | Same operation applied across multiple sections sequentially |
+
+**Default mode:** `interactive`
+
+**Mode inference:** "quickly" or "just fix" switches to `direct`. "step by step" or "逐步" confirms `interactive` (already default).
+
+## References
+
+### Required (always loaded)
+
+| File | Purpose |
+|------|---------|
+| `references/expression-patterns.md` | Academic expression patterns overview and module index |
+
+### Leaf Hints (loaded when needed)
+
+| File | When to Load |
+|------|--------------|
+| `references/expression-patterns/introduction-and-gap.md` | Polishing introduction or background content |
+| `references/expression-patterns/methods-and-data.md` | Polishing methods, data, or study area content |
+| `references/expression-patterns/results-and-discussion.md` | Polishing results or discussion content |
+| `references/expression-patterns/conclusions-and-claims.md` | Polishing conclusion content |
+| `references/expression-patterns/geography-domain.md` | Content involves spatial, urban, or planning topics |
+| `references/anti-ai-patterns.md` | Polishing expression (Step 3) -- screen for AI-sounding phrases |
+| `references/journals/ceus.md` | Target journal is CEUS |
+
+### Loading Rules
+
+- Load expression patterns overview at start; select the appropriate leaf based on section type.
+- Load anti-AI patterns when polishing expression (Step 3).
+- Load journal template when target journal is specified.
+- Load `geography-domain.md` when spatial, urban, or planning content is detected.
+- If a reference file is missing, warn the user and proceed with reduced capability.
+
+## Ask Strategy
+
+**Before starting, ask about:**
+1. Target journal (if not already known)
+2. Which section to work on
+3. Preferred mode (if ambiguous from trigger)
+
+**Rules:**
+- Never ask more than 3 questions before producing initial output.
+- In `direct` mode, skip pre-questions if the user provided enough context.
+- In `batch` mode, skip per-item questions; apply settings from the first item.
+- Use Structured Interaction when available; fall back to plain-text questions otherwise.
+- See `skill-conventions.md > AskUserQuestion Enforcement` for full rules.
+
+## Workflow
+
+### Step 1: Collect Context
+
+- Determine input type (file path or pasted text).
+- Load required references (expression-patterns overview).
+- Identify target journal; load journal template if specified.
+- Read input content using the Read tool; extract key numbers, claims, and data points.
+- Locate example/reference papers if the user provides them (use Read tool for PDFs).
+- In `interactive` or `guided` mode: confirm scope with the user before proceeding.
+
+### Step 2: Structure & Logic Confirmation
+
+- Analyze section macro structure (e.g., Abstract = Background + Gap + Method + Results + Contribution).
+- Present a structure table for user confirmation.
+- Break content into sentences; assign a logic function to each.
+- Present the logic chain for user confirmation.
+- Checkpoint: user confirms structure and logic before expression work begins.
+- In `direct` mode: run structure and logic analysis internally, proceed to Step 3 automatically.
+
+### Step 3: Expression Polish & Consistency
+
+- Load the section-appropriate expression pattern leaf and anti-AI patterns.
+- For each sentence with expression issues, present 2-3 options via AskUserQuestion (`interactive` mode) or apply the best option automatically (`direct` mode).
+
+```
+AskUserQuestion({
+  question: "Which expression do you prefer for [sentence function]?",
+  options: [
+    { label: "[Expression A]", description: "[full sentence with expression A]" },
+    { label: "[Expression B]", description: "[full sentence with expression B]" },
+    { label: "[Expression C]", description: "[full sentence with expression C]" }
+  ]
+})
+```
+
+- **Reference paper consultation:** when the user questions professionalism, use the Read tool to load example papers and extract expression patterns.
+- **Journal style check:** apply journal-specific requirements from the loaded template.
+- **Repetition and coherence pass:** check for repeated expressions and missing transitions; suggest fixes.
+- **Cross-section consistency:** verify numbers, terminology, and claims across sections.
+
+### Step 4: Output
+
+- Generate highlights if the journal requires them (for CEUS, see `references/journals/ceus.md`).
+- Suggest a read-aloud final check to catch awkward phrasing.
+- Compile all confirmed content into the final version.
+- Present the final version with word count for user confirmation.
+- Write to `*_polished.md` after confirmation (or automatically in `batch` mode).
+- Report word count and any journal constraint notes.
+
+## Output Contract
+
+| Output | Format | Condition |
+|--------|--------|-----------|
+| `polished_text` | Markdown file (`*_polished.md`) or conversation output | Always produced |
+| `change_summary` | Markdown in session | Always produced |
+| Word count | Integer | Always reported |
+| Journal compliance notes | Bullet list | When a target journal is specified |
+
+## Edge Cases
+
+| Situation | Handling |
+|-----------|----------|
+| Unprofessional word flagged | Present 2-3 alternatives via AskUserQuestion; accept if user insists |
+| Section too long for single pass | Split into paragraph-level sub-passes; maintain cross-paragraph coherence |
+| No journal specified | Default to general academic style; note in output |
+| Mixed language input | Detect dominant language; ask user to confirm target language |
+| Reference paper provided as PDF | Use Read tool to load PDF; extract style patterns for expression matching |
+| Abrupt sentence transition | Provide transition options via AskUserQuestion |
+| Repetition detected between sections | Identify repeated content; suggest which occurrence to rephrase |
+| Logic structure needs modification | Return to Step 2 to re-confirm structure before continuing |
+
+## Fallbacks
+
+| Scenario | Fallback |
+|----------|----------|
+| Structured Interaction unavailable | Ask 1-3 plain-text questions covering highest-impact gaps; do not block workflow |
+| Reference file missing | Log the missing file, proceed with reduced capability, warn the user |
+| Target journal not specified | Ask once; if declined, use general academic style |
+| PDF reference paper unreadable | Ask user to paste relevant excerpts instead |
+
+---
+
+*Skill: paper-polish-workflow*
+*Conventions: references/skill-conventions.md*

package/references/anti-ai-patterns/sentence-patterns.md

@@ -0,0 +1,33 @@
+# Sentence-Level AI Patterns
+
+Category focus: sentence templates that overstate certainty, flatten nuance, or package claims in a generic AI-writing cadence.
+
+## High Risk
+
+| Problem expression | Replacement | Why risky |
+|--------------------|-------------|-----------|
+| This proves that [claim]. | This suggests that [claim]. | Overstates certainty in most empirical papers |
+| It is undeniable that [claim]. | The evidence indicates that [claim]. | Removes academic caution |
+| This paper completely solves [problem]. | This paper addresses [problem] by [approach]. | Unrealistic and inflated |
+| There is no doubt that [claim]. | The findings consistently point to [claim]. | Too absolute for academic writing |
+
+## Medium Risk
+
+| Problem expression | Replacement | Why risky |
+|--------------------|-------------|-----------|
+| It can be clearly seen that [claim]. | The results show that [claim]. | Adds rhetorical padding rather than evidence |
+| This highlights the importance of [point]. | This indicates the relevance of [point]. | Common AI-style discussion filler |
+| It should be emphasized that [point]. | [Point]. | Usually redundant when the point is already strong |
+| This demonstrates the superiority of [method]. | This method performs better than [baseline] on [metric]. | Better to state the actual comparison |
+
+## Optional
+
+| Problem expression | Replacement | Why risky |
+|--------------------|-------------|-----------|
+| It is worth noting that [point]. | [Point]. | Often removable without losing meaning |
+| It is important to note that [point]. | [Point]. | Similar filler pattern |
+| In this context, [claim]. | [Claim]. | Fine occasionally, but repetitive if overused |
+
+---
+
+*Module: sentence-patterns*

package/references/anti-ai-patterns/transitions-and-tone.md

@@ -0,0 +1,32 @@
+# Transition and Tone Patterns
+
+Category focus: overly smooth transitions, polished filler, and tonal choices that make text sound generic rather than authorial.
+
+## High Risk
+
+| Problem expression | Replacement | Why risky |
+|--------------------|-------------|-----------|
+| Moreover, it is worth noting that [point]. | Moreover, [point]. | Double-padding in one sentence |
+| In conclusion, it is clear that [claim]. | In conclusion, these findings suggest that [claim]. | Wraps conclusion in unnecessary certainty |
+| Taken as a whole, this underscores the critical importance of [point]. | Taken together, these findings indicate that [point]. | Sounds polished but vague |
+
+## Medium Risk
+
+| Problem expression | Replacement | Why risky |
+|--------------------|-------------|-----------|
+| Furthermore, [generic reinforcement]. | Use only if the logical addition is real | Repeated transition stacking is a common AI signature |
+| On the other hand, [point]. | In contrast, [point]. / omit | Often misused when there is no true opposition |
+| This aligns with the broader trend that [claim]. | This pattern is consistent with prior findings on [topic]. | Generic tone without citation-ready specificity |
+| From a broader perspective, [claim]. | At the city scale / policy level / neighborhood level, [claim]. | Abstract framing sounds synthetic |
+
+## Optional
+
+| Problem expression | Replacement | Why risky |
+|--------------------|-------------|-----------|
+| Overall, [claim]. | Omit or move to conclusion | Fine once, weak when repeated in many sections |
+| In essence, [claim]. | [Claim]. | Stylistic filler |
+| Importantly, [claim]. | [Claim]. / specify the importance | Acceptable if used sparingly and with evidence |
+
+---
+
+*Module: transitions-and-tone*

package/references/anti-ai-patterns/vocabulary.md

@@ -0,0 +1,36 @@
+# Vocabulary Inflation Patterns
+
+Category focus: high-frequency words and adjective choices that often make academic text sound over-produced or obviously model-generated.
+
+## High Risk
+
+| Problem expression | Replacement | Why risky |
+|--------------------|-------------|-----------|
+| groundbreaking | useful in practice | Promotional, exaggerated, and often unsupported |
+| revolutionary | substantial | Reads like marketing instead of academic evaluation |
+| transformative | meaningful | Often signals inflated contribution language |
+| unprecedented | not previously reported in this dataset / context | Vague unless explicitly evidenced |
+| crucially | importantly / directly | Often over-emphasizes ordinary transitions |
+
+## Medium Risk
+
+| Problem expression | Replacement | Why risky |
+|--------------------|-------------|-----------|
+| robust | stable / consistent / reliable under [condition] | Too generic without context |
+| comprehensive | broad / multi-part / dataset-wide | Can hide what is actually covered |
+| insightful | informative | Tends to sound subjective |
+| seamless | coherent / smooth | Overly polished for academic prose |
+| leverage | use / draw on | Common AI-era business-academic hybrid wording |
+
+## Optional
+
+| Problem expression | Replacement | Why risky |
+|--------------------|-------------|-----------|
+| notably | omit or specify the point directly | Fine occasionally, weak when repeated |
+| significantly | materially / by X% / statistically significantly | Needs evidence or precision |
+| effective | effective at [task] | Acceptable if tied to evidence |
+| advanced | specific technical descriptor | Too vague if used alone |
+
+---
+
+*Module: vocabulary*

package/references/anti-ai-patterns.md

@@ -0,0 +1,56 @@
+# Anti-AI Patterns
+
+A modular reference library for detecting and rewriting AI-sounding academic phrasing.
+
+This overview file should remain the stable anti-AI entrypoint so future Skills can load it first and then choose one pattern category when needed.
+
+---
+
+## How to Use This Library
+
+- Read this file when you need the risk model, category map, or a quick retrieval contract.
+- Load one leaf file when the rewrite target is narrow, such as vocabulary inflation or transition overuse.
+- Default retrieval format for downstream Skills is `Problem expression -> Replacement`.
+- Use the risk tier to decide how aggressively to rewrite.
+
+## Risk Tiers
+
+| Tier | Meaning | Suggested use |
+|------|---------|---------------|
+| High Risk | Strong AI trace or promotional phrasing | Rewrite by default unless domain language truly requires it |
+| Medium Risk | Common smoothing pattern that may sound over-produced | Rewrite when tone feels too polished or repetitive |
+| Optional | Pattern is not necessarily wrong, but can sound generic if overused | Rewrite selectively based on surrounding context |
+
+## Module Map
+
+- `references/anti-ai-patterns/vocabulary.md`
+- `references/anti-ai-patterns/sentence-patterns.md`
+- `references/anti-ai-patterns/transitions-and-tone.md`
+
+## Lightweight Retrieval Layer
+
+| Category | Problem expression | Replacement |
+|----------|--------------------|-------------|
+| Vocabulary inflation | `groundbreaking` | `useful in practice` |
+| Sentence overclaim | `This proves that ...` | `This suggests that ...` |
+| Transition over-smoothing | `Moreover, it is worth noting that ...` | `Additionally, ...` or direct statement |
+
+## Stable Heading Contract
+
+Each leaf file should keep these headings:
+
+- `## High Risk`
+- `## Medium Risk`
+- `## Optional`
+
+## Maintenance Rules
+
+- Keep each entry short enough for direct reuse inside a Skill.
+- Prefer replacements that preserve formal academic tone.
+- Add rationale only when it helps a later Skill explain the rewrite choice.
+- Keep category filenames stable once downstream Skills begin referencing them.
+
+---
+
+*Entry point: references/anti-ai-patterns.md*
+*Updated for modular loading: 2026-03-11*