[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7EBC93DN78P57YPSA7Y2Y84":3,"skill-same:01M7EBC93DN78P57YPSA7Y2Y84":68},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":14,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":29,"ownerDescription":30,"updatedAt":31,"install":32,"resources":45},"01M7EBC93DN78P57YPSA7Y2Y84","pre-submission-reviewer",null,"Runs a pre-submission review of a technical paper across five dimensions: macro logic, writing details, English grammar, LaTeX formatting, and figure quality. Uses a reviewer-style severity taxonomy (CRITICAL \u002F MAJOR \u002F MINOR) and flags banned AI-tone vocabulary and em-dash misuse. Use when the user asks to 'review this paper', 'audit before submission', 'check the draft', 'find issues', 'proofread', or within one week of a submission deadline.","zms618\u002Fno-heartburn-academic-skills","zms618","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F105599679?v=4","01M7EBA1W5DCCS8A3KMY0A0R28","Cheng He","01M7EBA1XAKYE8NVWVS162527D",0,"CC-BY-4.0","https:\u002F\u002Fgithub.com\u002Fzms618\u002Fno-heartburn-academic-skills\u002Fblob\u002Fmain\u002Fskills\u002Fai-mentor\u002Fskills\u002Fpre-submission-reviewer\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: pre-submission-reviewer\ndescription: >-\n  Runs a pre-submission review of a technical paper across five\n  dimensions: macro logic, writing details, English grammar, LaTeX\n  formatting, and figure quality. Uses a reviewer-style severity\n  taxonomy (CRITICAL \u002F MAJOR \u002F MINOR) and flags banned AI-tone\n  vocabulary and em-dash misuse. Use when the user asks to 'review\n  this paper', 'audit before submission', 'check the draft', 'find\n  issues', 'proofread', or within one week of a submission\n  deadline.\nlicense: CC-BY-4.0\n---\n\n# Pre-Submission Reviewer\n\n## Overview\n\nThree to five days before a submission deadline is the window where\na careful external review pays off most. This skill takes a full\npaper or key sections and produces a structured review across five\ndimensions, each with severity-tagged findings and concrete\nrewrite suggestions. It enforces the mechanical rules from the\nwriting-checklist section (no em-dashes, no banned AI-tone\nvocabulary, leading text per paragraph, topic-sentence discipline,\ncitation-format uniformity) and surfaces the patterns that\nnon-native English-speaking authors most commonly violate\n(articles, subject-verb agreement, tense consistency, which versus\nthat, Chinglish phrasing).\n\nThe output is not a rewrite. It is a prioritised list of findings\nwith severity tags; the author decides which to fix. CRITICAL items\nshould block submission until addressed.\n\n## When to use this skill\n\n- Three to five days before a submission deadline.\n- The user asks to 'review this paper', 'audit before submission',\n  'check the draft', 'find issues', 'proofread'.\n- After a camera-ready revision, before sending the final version.\n- After any major rewrite (rebuttal responses, Section 3 overhaul).\n- When the user suspects AI-tone contamination in a section.\n\n## When NOT to use this skill\n\n- The paper is still being structured. Use `tech-paper-template`,\n  `intro-drafter`, or `benchmark-paper-template` (separate plugin) first.\n- The user wants structural advice rather than review. Use the\n  drafting skills instead.\n\n## Core procedure\n\n### Step 0: paradigm and venue fit\n\nBefore the dimensional review, settle two things.\n\n**Paradigm**, judged by research method, because what counts as a\nsevere problem differs by paradigm:\n\n- **STEM \u002F technical** (CS, engineering, materials, chemistry): weight\n  the Introduction chain, contribution-to-section mapping, baseline\n  coverage, ablation-based attribution, figure quality. Benchmark\n  papers additionally get coverage, reproducibility, and\n  contamination checks; chemistry and materials get characterization\n  completeness, purity, controls, replicates.\n- **Humanities** (literature, history, philosophy, discourse\n  analysis): weight whether the thesis is explicit and defensible,\n  core concepts pinned down and stable across sections, the\n  literature genuinely engaged, and the material (texts, archives,\n  cases) able to carry the claims. Do **not** demand baselines or\n  ablations here; check instead whether sub-arguments build on each\n  other rather than sit in parallel, and whether conclusions outrun\n  the material.\n- **Empirical social science**: weight operationalized research\n  questions, sampling and data-source justification, statistics\n  matched to data types, conclusions bounded by the sample. For\n  theory papers, \"experiments\" reads as \"proofs\": check the proofs,\n  and never call the evidence thin merely because there is no results\n  table.\n- **Finance \u002F economics**: weight identification credibility,\n  endogeneity handling, robustness checks (their absence is a\n  first-round flag), and economic versus merely statistical\n  significance.\n- **Law**: weight the accuracy of statutes, case numbers, and\n  holdings (an invented or wrong citation is rejection-level), the\n  clarity of the interpretive approach, and whether comparative\n  arguments state their scope.\n\nWhen the paradigm is unclear, ask the author before reviewing with\nthe wrong ruler.\n\n**Venue fit**: if the stated target venue's scope visibly mismatches\nthe paper's topic or contribution type, that is a real rejection\nrisk, not a taste note. Flag it as a finding and suggest two or three\nbetter-fitting venues.\n\n### Step 1: Dimension 1 Macro logic review\n\nSee: references\u002Flogic-and-structure.md for the Logic First rule,\nSelf-contained rule, Leading Text rule, and Running Example rule.\n\nCheck:\n\n- Introduction flowchart is intact (Background, Limitations, Goal\n  or Key Idea, Challenges, Methodology, Contributions).\n- Contributions map one-to-one with methodology modules and with\n  section numbers.\n- Experiments validate the paper's main claims, not tangential\n  ones.\n- Related Work covers the necessary prior art.\n- Running example is consistent across Introduction, Methodology,\n  Experiments.\n- The headline result's attribution is isolated: an ablation\n  separates the core mechanism from peripheral factors (a routing\n  step, post-processing, a stronger base model, favorable samples).\n  No such ablation: flag \"attribution unverified\" as MAJOR.\n- Claims match their evidence: \"solves\" is stronger than most papers\n  earn (usually \"improves\"); \"state-of-the-art\" needs the benchmark\n  and conditions; \"we are the first\" gets checked or flagged.\n\nEvery break in the chain is CRITICAL.\n\n**Retrieval-grounded checks** (when the environment has a\nliterature-search capability: a scholarly tool, web search over\nscholarly indexes, or shell access to public APIs):\n\n1. **Novelty verification**: extract two or three keyword groups from\n   the paper's core method and problem setting, retrieve, and identify\n   the three to five closest published works. If Related Work already\n   covers them, the paper's positioning stands; a highly relevant\n   uncovered work is a MAJOR finding (\"missed X, Author et al.,\n   Year\"). Compare on difference axes; a similar title alone proves\n   nothing.\n2. **Citation completeness**: retrieve the field's recent\n   representative works and its canonical ones, and compare against\n   the reference list; a missing canonical baseline, founding paper,\n   or recent survey is MAJOR.\n\nRetrieval results support metadata-level judgments only; never quote\nnumbers or method details from search snippets. Without any retrieval\ncapability, skip these two checks and say so in the summary.\n\n### Step 2: Dimension 2 Writing details review\n\nSee: references\u002Flogic-and-structure.md for paragraph-level rules.\n\nCheck:\n\n- Every paragraph has a topic sentence.\n- Paragraphs transition smoothly; no orphan paragraphs.\n- Paragraphs are not over 10 lines; split if so.\n- No repeated or redundant passages.\n- Abstract covers problem, method, result.\n\n### Step 3: Dimension 3 English grammar review\n\nSee: references\u002Fgrammar-rules.md for the canonical list of errors\ncommon to non-native English authors, with corrections and\nexamples.\n\nCheck the usual suspects:\n\n- Article use (a, an, the).\n- Subject-verb agreement (third-person singular).\n- Tense consistency (Related Work past, method present).\n- Passive-voice overuse.\n- Which versus that.\n- Sentence length; split long sentences at \"Specifically,\".\n- Chinglish patterns.\n\n### Step 4: Dimension 4 LaTeX format review\n\nSee: references\u002Flatex-rules.md for the canonical list of LaTeX-\nspecific issues.\n\nCheck:\n\n- Equation numbering contiguous; every numbered equation\n  referenced.\n- Figures and tables have captions; captions are detailed.\n- Citations use the correct command and the non-breaking tilde\n  (for example, `ResNet~\\cite{X}`, never `ResNet\\cite{X}`).\n- Labels use underscores, not spaces or hyphens.\n- Vector figure format; no raster.\n- Page-limit compliance.\n\n### Step 5: Dimension 5 Figure quality review\n\nSee: references\u002Fforbidden-patterns.md for chartjunk patterns and\nthe full figure-quality checklist.\n\nFor each figure:\n\n- Vector format.\n- Font size large enough post-scaling.\n- Colour-blind-safe palette; dual encoding.\n- Self-contained caption with a finding in the first sentence.\n- No chartjunk.\n- Motivated example is concrete and failure-revealing.\n- Solution overview has labels matching section titles.\n\n### Step 6: Banned-vocabulary and em-dash scan\n\nSee: references\u002Fforbidden-patterns.md for the banned-word list.\n\nScan the full paper for:\n\n- Em-dashes used as sentence connectors (banned; project rule).\n- AI-tone words: innovative, pioneering, revolutionary paradigm,\n  transformative framework, superior, surpass, excel, remarkable,\n  unprecedented, breakthrough performance, general-purpose, is\n  capable of, notably, yet, yielding, at its essence, encompass,\n  differentiate, reveal, underscore, pave the way for, highlight\n  the potential of, profound challenges, stems from, rigid,\n  impede.\n\nFlag each occurrence with a severity tag. Em-dashes are MAJOR by\ndefault; banned AI-tone words are MAJOR if they appear three or\nmore times.\n\n### Step 7: Section-by-section review\n\nSee: references\u002Fsection-guides.md for the per-section writing\nguides for Abstract, Introduction, Problem Formulation, Framework\nor Method, Experiments, Related Work, and Conclusion.\n\nFor each section, check that the section's content matches the\nguide's canonical structure (for example, Abstract's five-sentence\nformula: what, why, challenges, how, results).\n\n### Step 8: Integrity gate\n\nRun the checks in the Integrity gate section below.\n\n### Step 9: Output\n\nEmit the review in the Output format below.\n\n## Severity taxonomy\n\n- **CRITICAL**: blocks submission. Example: contributions do not\n  map to sections; introduction flowchart broken; no real-world\n  running example; raster figure in final draft; missing key\n  baseline; page-limit violation.\n- **MAJOR**: reviewers will flag in first round. Example:\n  topic-sentence absent from 3+ paragraphs; em-dash in 5+ places;\n  banned AI-tone word in 3+ places; Table 1 comparison missing;\n  chart type mismatched with data.\n- **MINOR**: polish. Example: two long sentences that could be\n  split; default Matplotlib styling; single article error.\n\nSeverity honesty cuts both ways. A review that lists a dozen MINOR\nitems while missing the one rejection-level flaw sends the author to\nsubmission with false confidence; a review that inflates taste issues\ninto CRITICAL destroys trust. The overall recommendation must match\nthe findings: any unresolved CRITICAL forbids \"ready to submit\", and\na near-ready verdict requires zero CRITICAL and at most two MAJOR.\n\n## Integrity gate\n\nEach bullet is tagged [inspection] (LLM verifies from the paper\ntext) or [attestation] (LLM runs the procedure and states it has\ndone so; user remains responsible for confirming completeness).\n\nBefore emitting the review:\n\n1. **[inspection]** Every finding quotes specific text (sentence,\n   phrase, figure name); no \"the Introduction is unclear\" without\n   a quoted line.\n2. **[inspection]** Every CRITICAL finding has a concrete fix\n   suggestion, not \"rewrite entirely\".\n3. **[inspection]** No fabricated quotes: only text actually\n   present in the submitted material.\n4. **[inspection]** Severity assignments follow the taxonomy;\n   nothing is marked CRITICAL for taste reasons.\n5. **[inspection]** Dimension 3 (grammar) findings cite the\n   specific grammar rule from `references\u002Fgrammar-rules.md`.\n6. **[attestation]** Dimension 6 banned-vocabulary scan is run in\n   full on the entire paper, not sampled. The skill attests the\n   full scan; if the paper is extremely long, the skill states it\n   chunked the input and describes the chunking strategy.\n7. **[inspection]** Final score matches the CRITICAL + MAJOR\n   count; a score of 9 or 10 requires zero CRITICAL and at most\n   two MAJOR items.\n\nIf any [inspection] check fails, mark the output as \"needs user\nattention\". For [attestation] bullets, the skill states the scope\nof its scan and the user confirms completeness.\n\nRun the gate silently. Do not print a per-gate pass or fail report;\na failure surfaces as a concrete finding in the affected dimension,\nand the delivered review stays free of internal checking rituals.\n\n## Output format\n\n### Summary\n- CRITICAL: \u003Cn>\n- MAJOR: \u003Cm>\n- MINOR: \u003Ck>\n- Top three fixes first: ...\n\n### Dimension 1: Macro logic\n| # | Finding | Severity | Suggested fix |\n|---|---|---|---|\n| 1 | \u003Cquoted text> | CRITICAL or MAJOR or MINOR | \u003Cfix> |\n\n### Dimension 2: Writing details\n\u003Csame table shape>\n\n### Dimension 3: English grammar\n\u003Csame table shape, citing grammar-rule ID>\n\n### Dimension 4: LaTeX format\n\u003Csame table shape>\n\n### Dimension 5: Figure quality\n\u003Csame table shape>\n\n### Banned-vocabulary and em-dash scan\n\u003Clist with line references>\n\n### Final score (1-10)\n\u003Cscore>\n\n### Submission recommendation\n- \u003CReady to submit | Needs 1-2 days more work | Needs major revision before submission>\n","# Pre-Submission Reviewer\n\n## Overview\n\nThree to five days before a submission deadline is the window where\na careful external review pays off most. This skill takes a full\npaper or key sections and produces a structured review across five\ndimensions, each with severity-tagged findings and concrete\nrewrite suggestions. It enforces the mechanical rules from the\nwriting-checklist section (no em-dashes, no banned AI-tone\nvocabulary, leading text per paragraph, topic-sentence discipline,\ncitation-format uniformity) and surfaces the patterns that\nnon-native English-speaking authors most commonly violate\n(articles, subject-verb agreement, tense consistency, which versus\nthat, Chinglish phrasing).\n\nThe output is not a rewrite. It is a prioritised list of findings\nwith severity tags; the author decides which to fix. CRITICAL items\nshould block submission until addressed.\n\n## When to use this skill\n\n- Three to five days before a submission deadline.\n- The user asks to 'review this paper', 'audit before submission',\n  'check the draft', 'find issues', 'proofread'.\n- After a camera-ready revision, before sending the final version.\n- After any major rewrite (rebuttal responses, Section 3 overhaul).\n- When the user suspects AI-tone contamination in a section.\n\n## When NOT to use this skill\n\n- The paper is still being structured. Use `tech-paper-template`,\n  `intro-drafter`, or `benchmark-paper-template` (separate plugin) first.\n- The user wants structural advice rather than review. Use the\n  drafting skills instead.\n\n## Core procedure\n\n### Step 0: paradigm and venue fit\n\nBefore the dimensional review, settle two things.\n\n**Paradigm**, judged by research method, because what counts as a\nsevere problem differs by paradigm:\n\n- **STEM \u002F technical** (CS, engineering, materials, chemistry): weight\n  the Introduction chain, contribution-to-section mapping, baseline\n  coverage, ablation-based attribution, figure quality. Benchmark\n  papers additionally get coverage, reproducibility, and\n  contamination checks; chemistry and materials get characterization\n  completeness, purity, controls, replicates.\n- **Humanities** (literature, history, philosophy, discourse\n  analysis): weight whether the thesis is explicit and defensible,\n  core concepts pinned down and stable across sections, the\n  literature genuinely engaged, and the material (texts, archives,\n  cases) able to carry the claims. Do **not** demand baselines or\n  ablations here; check instead whether sub-arguments build on each\n  other rather than sit in parallel, and whether conclusions outrun\n  the material.\n- **Empirical social science**: weight operationalized research\n  questions, sampling and data-source justification, statistics\n  matched to data types, conclusions bounded by the sample. For\n  theory papers, \"experiments\" reads as \"proofs\": check the proofs,\n  and never call the evidence thin merely because there is no results\n  table.\n- **Finance \u002F economics**: weight identification credibility,\n  endogeneity handling, robustness checks (their absence is a\n  first-round flag), and economic versus merely statistical\n  significance.\n- **Law**: weight the accuracy of statutes, case numbers, and\n  holdings (an invented or wrong citation is rejection-level), the\n  clarity of the interpretive approach, and whether comparative\n  arguments state their scope.\n\nWhen the paradigm is unclear, ask the author before reviewing with\nthe wrong ruler.\n\n**Venue fit**: if the stated target venue's scope visibly mismatches\nthe paper's topic or contribution type, that is a real rejection\nrisk, not a taste note. Flag it as a finding and suggest two or three\nbetter-fitting venues.\n\n### Step 1: Dimension 1 Macro logic review\n\nSee: references\u002Flogic-and-structure.md for the Logic First rule,\nSelf-contained rule, Leading Text rule, and Running Example rule.\n\nCheck:\n\n- Introduction flowchart is intact (Background, Limitations, Goal\n  or Key Idea, Challenges, Methodology, Contributions).\n- Contributions map one-to-one with methodology modules and with\n  section numbers.\n- Experiments validate the paper's main claims, not tangential\n  ones.\n- Related Work covers the necessary prior art.\n- Running example is consistent across Introduction, Methodology,\n  Experiments.\n- The headline result's attribution is isolated: an ablation\n  separates the core mechanism from peripheral factors (a routing\n  step, post-processing, a stronger base model, favorable samples).\n  No such ablation: flag \"attribution unverified\" as MAJOR.\n- Claims match their evidence: \"solves\" is stronger than most papers\n  earn (usually \"improves\"); \"state-of-the-art\" needs the benchmark\n  and conditions; \"we are the first\" gets checked or flagged.\n\nEvery break in the chain is CRITICAL.\n\n**Retrieval-grounded checks** (when the environment has a\nliterature-search capability: a scholarly tool, web search over\nscholarly indexes, or shell access to public APIs):\n\n1. **Novelty verification**: extract two or three keyword groups from\n   the paper's core method and problem setting, retrieve, and identify\n   the three to five closest published works. If Related Work already\n   covers them, the paper's positioning stands; a highly relevant\n   uncovered work is a MAJOR finding (\"missed X, Author et al.,\n   Year\"). Compare on difference axes; a similar title alone proves\n   nothing.\n2. **Citation completeness**: retrieve the field's recent\n   representative works and its canonical ones, and compare against\n   the reference list; a missing canonical baseline, founding paper,\n   or recent survey is MAJOR.\n\nRetrieval results support metadata-level judgments only; never quote\nnumbers or method details from search snippets. Without any retrieval\ncapability, skip these two checks and say so in the summary.\n\n### Step 2: Dimension 2 Writing details review\n\nSee: references\u002Flogic-and-structure.md for paragraph-level rules.\n\nCheck:\n\n- Every paragraph has a topic sentence.\n- Paragraphs transition smoothly; no orphan paragraphs.\n- Paragraphs are not over 10 lines; split if so.\n- No repeated or redundant passages.\n- Abstract covers problem, method, result.\n\n### Step 3: Dimension 3 English grammar review\n\nSee: references\u002Fgrammar-rules.md for the canonical list of errors\ncommon to non-native English authors, with corrections and\nexamples.\n\nCheck the usual suspects:\n\n- Article use (a, an, the).\n- Subject-verb agreement (third-person singular).\n- Tense consistency (Related Work past, method present).\n- Passive-voice overuse.\n- Which versus that.\n- Sentence length; split long sentences at \"Specifically,\".\n- Chinglish patterns.\n\n### Step 4: Dimension 4 LaTeX format review\n\nSee: references\u002Flatex-rules.md for the canonical list of LaTeX-\nspecific issues.\n\nCheck:\n\n- Equation numbering contiguous; every numbered equation\n  referenced.\n- Figures and tables have captions; captions are detailed.\n- Citations use the correct command and the non-breaking tilde\n  (for example, `ResNet~\\cite{X}`, never `ResNet\\cite{X}`).\n- Labels use underscores, not spaces or hyphens.\n- Vector figure format; no raster.\n- Page-limit compliance.\n\n### Step 5: Dimension 5 Figure quality review\n\nSee: references\u002Fforbidden-patterns.md for chartjunk patterns and\nthe full figure-quality checklist.\n\nFor each figure:\n\n- Vector format.\n- Font size large enough post-scaling.\n- Colour-blind-safe palette; dual encoding.\n- Self-contained caption with a finding in the first sentence.\n- No chartjunk.\n- Motivated example is concrete and failure-revealing.\n- Solution overview has labels matching section titles.\n\n### Step 6: Banned-vocabulary and em-dash scan\n\nSee: references\u002Fforbidden-patterns.md for the banned-word list.\n\nScan the full paper for:\n\n- Em-dashes used as sentence connectors (banned; project rule).\n- AI-tone words: innovative, pioneering, revolutionary paradigm,\n  transformative framework, superior, surpass, excel, remarkable,\n  unprecedented, breakthrough performance, general-purpose, is\n  capable of, notably, yet, yielding, at its essence, encompass,\n  differentiate, reveal, underscore, pave the way for, highlight\n  the potential of, profound challenges, stems from, rigid,\n  impede.\n\nFlag each occurrence with a severity tag. Em-dashes are MAJOR by\ndefault; banned AI-tone words are MAJOR if they appear three or\nmore times.\n\n### Step 7: Section-by-section review\n\nSee: references\u002Fsection-guides.md for the per-section writing\nguides for Abstract, Introduction, Problem Formulation, Framework\nor Method, Experiments, Related Work, and Conclusion.\n\nFor each section, check that the section's content matches the\nguide's canonical structure (for example, Abstract's five-sentence\nformula: what, why, challenges, how, results).\n\n### Step 8: Integrity gate\n\nRun the checks in the Integrity gate section below.\n\n### Step 9: Output\n\nEmit the review in the Output format below.\n\n## Severity taxonomy\n\n- **CRITICAL**: blocks submission. Example: contributions do not\n  map to sections; introduction flowchart broken; no real-world\n  running example; raster figure in final draft; missing key\n  baseline; page-limit violation.\n- **MAJOR**: reviewers will flag in first round. Example:\n  topic-sentence absent from 3+ paragraphs; em-dash in 5+ places;\n  banned AI-tone word in 3+ places; Table 1 comparison missing;\n  chart type mismatched with data.\n- **MINOR**: polish. Example: two long sentences that could be\n  split; default Matplotlib styling; single article error.\n\nSeverity honesty cuts both ways. A review that lists a dozen MINOR\nitems while missing the one rejection-level flaw sends the author to\nsubmission with false confidence; a review that inflates taste issues\ninto CRITICAL destroys trust. The overall recommendation must match\nthe findings: any unresolved CRITICAL forbids \"ready to submit\", and\na near-ready verdict requires zero CRITICAL and at most two MAJOR.\n\n## Integrity gate\n\nEach bullet is tagged [inspection] (LLM verifies from the paper\ntext) or [attestation] (LLM runs the procedure and states it has\ndone so; user remains responsible for confirming completeness).\n\nBefore emitting the review:\n\n1. **[inspection]** Every finding quotes specific text (sentence,\n   phrase, figure name); no \"the Introduction is unclear\" without\n   a quoted line.\n2. **[inspection]** Every CRITICAL finding has a concrete fix\n   suggestion, not \"rewrite entirely\".\n3. **[inspection]** No fabricated quotes: only text actually\n   present in the submitted material.\n4. **[inspection]** Severity assignments follow the taxonomy;\n   nothing is marked CRITICAL for taste reasons.\n5. **[inspection]** Dimension 3 (grammar) findings cite the\n   specific grammar rule from `references\u002Fgrammar-rules.md`.\n6. **[attestation]** Dimension 6 banned-vocabulary scan is run in\n   full on the entire paper, not sampled. The skill attests the\n   full scan; if the paper is extremely long, the skill states it\n   chunked the input and describes the chunking strategy.\n7. **[inspection]** Final score matches the CRITICAL + MAJOR\n   count; a score of 9 or 10 requires zero CRITICAL and at most\n   two MAJOR items.\n\nIf any [inspection] check fails, mark the output as \"needs user\nattention\". For [attestation] bullets, the skill states the scope\nof its scan and the user confirms completeness.\n\nRun the gate silently. Do not print a per-gate pass or fail report;\na failure surfaces as a concrete finding in the affected dimension,\nand the delivered review stays free of internal checking rituals.\n\n## Output format\n\n### Summary\n- CRITICAL: \u003Cn>\n- MAJOR: \u003Cm>\n- MINOR: \u003Ck>\n- Top three fixes first: ...\n\n### Dimension 1: Macro logic\n| # | Finding | Severity | Suggested fix |\n|---|---|---|---|\n| 1 | \u003Cquoted text> | CRITICAL or MAJOR or MINOR | \u003Cfix> |\n\n### Dimension 2: Writing details\n\u003Csame table shape>\n\n### Dimension 3: English grammar\n\u003Csame table shape, citing grammar-rule ID>\n\n### Dimension 4: LaTeX format\n\u003Csame table shape>\n\n### Dimension 5: Figure quality\n\u003Csame table shape>\n\n### Banned-vocabulary and em-dash scan\n\u003Clist with line references>\n\n### Final score (1-10)\n\u003Cscore>\n\n### Submission recommendation\n- \u003CReady to submit | Needs 1-2 days more work | Needs major revision before submission>\n",{"name":5,"license":15,"description":7},"main","e6390f546a3cc4731cf8bd03a19a20ca8334553b","skills\u002Fai-mentor\u002Fskills\u002Fpre-submission-reviewer","https:\u002F\u002Fgithub.com\u002Fzms618\u002Fno-heartburn-academic-skills","https:\u002F\u002Fgithub.com\u002Fzms618","https:\u002F\u002Fzms618.github.io\u002F","M.S. Student @ Fuzhou University | Computational Pathology & Medical Vision","2026-10-13 13:38:55",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":44},[34,37,40],{"tool":35,"command":36},"npx","npx skills add zms618\u002Fno-heartburn-academic-skills --skill pre-submission-reviewer",{"tool":38,"command":39},"bunx","bunx skills add zms618\u002Fno-heartburn-academic-skills --skill pre-submission-reviewer",{"tool":41,"command":42},"pnpm","pnpm dlx skills add zms618\u002Fno-heartburn-academic-skills --skill pre-submission-reviewer","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fzms618\u002Fno-heartburn-academic-skills\u002Fblob\u002Fmain\u002Fskills\u002Fai-mentor\u002Fskills\u002Fpre-submission-reviewer\u002FSKILL.md",true,[46,52,56,60,64],{"path":47,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":50,"sourceHtmlUrl":51},"references\u002Fforbidden-patterns.md","REFERENCE","参考文档",4801,"https:\u002F\u002Fgithub.com\u002Fzms618\u002Fno-heartburn-academic-skills\u002Fblob\u002Fmain\u002Fskills\u002Fai-mentor\u002Fskills\u002Fpre-submission-reviewer\u002Freferences\u002Fforbidden-patterns.md",{"path":53,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":54,"sourceHtmlUrl":55},"references\u002Fgrammar-rules.md",4767,"https:\u002F\u002Fgithub.com\u002Fzms618\u002Fno-heartburn-academic-skills\u002Fblob\u002Fmain\u002Fskills\u002Fai-mentor\u002Fskills\u002Fpre-submission-reviewer\u002Freferences\u002Fgrammar-rules.md",{"path":57,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":58,"sourceHtmlUrl":59},"references\u002Flatex-rules.md",4439,"https:\u002F\u002Fgithub.com\u002Fzms618\u002Fno-heartburn-academic-skills\u002Fblob\u002Fmain\u002Fskills\u002Fai-mentor\u002Fskills\u002Fpre-submission-reviewer\u002Freferences\u002Flatex-rules.md",{"path":61,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":62,"sourceHtmlUrl":63},"references\u002Flogic-and-structure.md",4663,"https:\u002F\u002Fgithub.com\u002Fzms618\u002Fno-heartburn-academic-skills\u002Fblob\u002Fmain\u002Fskills\u002Fai-mentor\u002Fskills\u002Fpre-submission-reviewer\u002Freferences\u002Flogic-and-structure.md",{"path":65,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":66,"sourceHtmlUrl":67},"references\u002Fsection-guides.md",5334,"https:\u002F\u002Fgithub.com\u002Fzms618\u002Fno-heartburn-academic-skills\u002Fblob\u002Fmain\u002Fskills\u002Fai-mentor\u002Fskills\u002Fpre-submission-reviewer\u002Freferences\u002Fsection-guides.md",[69,76,82,89,96,103],{"skillNo":70,"name":71,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":72,"licenseFileUrl":6,"sourceHtmlUrl":73,"categories":74,"keywords":75,"publishedAt":19},"01M7EBH8ZAKH92EA8GG2A4Z8B4","research-review-board","MIT","https:\u002F\u002Fgithub.com\u002Fzms618\u002Fno-heartburn-academic-skills\u002Fblob\u002Fmain\u002Fskills\u002Fresearch-idea-discovery\u002Fskills\u002Fresearch-review-board\u002FSKILL.md",[],[],{"skillNo":77,"name":78,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":72,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":19},"01M7EBH5BYYHV4EVM1YDP0JXWW","research-proposal-generator","https:\u002F\u002Fgithub.com\u002Fzms618\u002Fno-heartburn-academic-skills\u002Fblob\u002Fmain\u002Fskills\u002Fresearch-idea-discovery\u002Fskills\u002Fresearch-proposal-generator\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":72,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":19},"01M7EBH1PX2Q7R48QY45Q94D3T","research-problem-validation","Validate research problems before generating ideas.","https:\u002F\u002Fgithub.com\u002Fzms618\u002Fno-heartburn-academic-skills\u002Fblob\u002Fmain\u002Fskills\u002Fresearch-idea-discovery\u002Fskills\u002Fresearch-problem-validation\u002FSKILL.md",[],[],{"skillNo":90,"name":91,"nameZh":6,"description":92,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":72,"licenseFileUrl":6,"sourceHtmlUrl":93,"categories":94,"keywords":95,"publishedAt":19},"01M7EBGVT2CFZNEMJSBXWYX751","research-problem-scout","Use at the beginning of scientific idea discovery to find real, important, falsifiable research failures from actual sources; reject weak motivations before architectures and novelty-hunting. Includes low-cost problem-evidence scouting and a hard Motivation-First Gate.","https:\u002F\u002Fgithub.com\u002Fzms618\u002Fno-heartburn-academic-skills\u002Fblob\u002Fmain\u002Fskills\u002Fresearch-idea-discovery\u002Fskills\u002Fresearch-problem-scout\u002FSKILL.md",[],[],{"skillNo":97,"name":98,"nameZh":6,"description":99,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":72,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":102,"publishedAt":19},"01M7EBGQ1WQ7ZQ8DCDYXVBT1W9","research-pilot-designer","Use when user requests falsification pilot experiment design hypothesis testing research planning. Works within the general Research Idea Discovery plugin and any user-selected field.","https:\u002F\u002Fgithub.com\u002Fzms618\u002Fno-heartburn-academic-skills\u002Fblob\u002Fmain\u002Fskills\u002Fresearch-idea-discovery\u002Fskills\u002Fresearch-pilot-designer\u002FSKILL.md",[],[],{"skillNo":104,"name":105,"nameZh":6,"description":106,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":72,"licenseFileUrl":6,"sourceHtmlUrl":107,"categories":108,"keywords":109,"publishedAt":19},"01M7EBGH5KFCBRGNQXS0TYHXE4","research-paper-anchor-scout","Critically read a small adversarial group of anchor papers to discover real, falsifiable research problems from failures, successful-method mechanisms, contradictions and deployment assumptions; reuse verified assets without blindly modifying the previous network.","https:\u002F\u002Fgithub.com\u002Fzms618\u002Fno-heartburn-academic-skills\u002Fblob\u002Fmain\u002Fskills\u002Fresearch-idea-discovery\u002Fskills\u002Fresearch-paper-anchor-scout\u002FSKILL.md",[],[]]