[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6FJWHYC5YN6B8PJFCWDBAXN":3,"skill-same:01M6FJWHYC5YN6B8PJFCWDBAXN":53},{"skillNo":4,"name":5,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":22,"publishedAt":26,"linkOnly":27,"markdownContent":28,"markdownBody":28,"frontMatter":29,"refName":30,"commitSha":31,"skillDirectory":32,"forksCount":33,"repositoryHtmlUrl":34,"repositoryHomepageUrl":35,"ownerHtmlUrl":36,"ownerBlogUrl":6,"ownerDescription":37,"updatedAt":38,"install":39,"resources":52},"01M6FJWHYC5YN6B8PJFCWDBAXN","paper-plan",null,"ComeOnOliver\u002Fskillshub","ComeOnOliver","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F43215937?v=4","01M6E5JNFECJAAVFE0W7SMWWD2","Haoran Shu","01M6E5JNNDQS7AJZDYFQ8TPZCK",67,"MIT","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fwanshuiyin\u002FAuto-claude-code-research-in-sleep\u002Fpaper-plan\u002FSKILL.md",[17],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000021","资料检索与研究","SCENARIO","使用场景",[23,24,25],"plan","paper","paper plan","2026-11-10 10:12:30",false,"# Paper Plan: From Review Conclusions to Paper Outline\n\nGenerate a structured, section-by-section paper outline from: **$ARGUMENTS**\n\n## Constants\n\n- **REVIEWER_MODEL = `gpt-5.4`** — Model used via Codex MCP for outline review. Must be an OpenAI model.\n- **TARGET_VENUE = `ICLR`** — Default venue. User can override (e.g., `\u002Fpaper-plan \"topic\" — venue: NeurIPS`). Supported: `ICLR`, `NeurIPS`, `ICML`.\n- **MAX_PAGES** — Main body page limit, measured from first page to end of Conclusion section (excluding references, appendix, and acknowledgements). ICLR=9, NeurIPS=9, ICML=8.\n\n## Inputs\n\nThe skill expects one or more of these in the project directory:\n\n1. **NARRATIVE_REPORT.md** or **STORY.md** — research narrative with claims and evidence\n2. **GPT54_AUTO_REVIEW.md** — auto-review loop conclusions\n3. **Experiment results** — JSON files in `figures\u002F`, screen logs, tables\n4. **IDEA_REPORT.md** — from idea-discovery pipeline (if applicable)\n\nIf none exist, ask the user to describe the paper's contribution in 3-5 sentences.\n\n## Workflow\n\n### Step 1: Extract Claims and Evidence\n\nRead all available narrative documents and extract:\n\n1. **Core claims** (3-5 main contributions)\n2. **Evidence** for each claim (which experiments, which metrics, which figures)\n3. **Known weaknesses** (from reviewer feedback)\n4. **Suggested framing** (from review conclusions)\n\nBuild a **Claims-Evidence Matrix**:\n\n```markdown\n| Claim | Evidence | Status | Section |\n|-------|----------|--------|---------|\n| [claim 1] | [exp A, metric B] | Supported | §3.2 |\n| [claim 2] | [exp C] | Partially supported | §4.1 |\n```\n\n### Step 2: Determine Paper Type and Structure\n\nBased on TARGET_VENUE and paper content, classify and select structure.\n\n**IMPORTANT**: The section count is FLEXIBLE (5-8 sections). Choose what fits the content best. The templates below are starting points, not rigid constraints.\n\n**Empirical\u002FDiagnostic paper:**\n```\n1. Introduction (1.5 pages)\n2. Related Work (1 page)\n3. Method \u002F Setup (1.5 pages)\n4. Experiments (3 pages)\n5. Analysis \u002F Discussion (1 page)\n6. Conclusion (0.5 pages)\n```\n\n**Theory + Experiments paper:**\n```\n1. Introduction (1.5 pages)\n2. Related Work (1 page)\n3. Preliminaries & Modeling (1.5 pages)\n4. Experiments (1.5 pages)\n5. Theory Part A (1.5 pages)\n6. Theory Part B (1.5 pages)\n7. Conclusion (0.5 pages)\n— Total: 9 pages\n```\nTheory papers often need 7 sections (splitting theory into estimation + optimization, or setup + analysis). The total page budget MUST sum to MAX_PAGES.\n\nTheory papers should:\n- Include **proof sketch** locations (not just theorem statements)\n- Plan a **comparison table** of prior theoretical bounds vs. this paper's bounds\n- Identify which proofs go in appendix vs. main body\n\n**Method paper:**\n```\n1. Introduction (1.5 pages)\n2. Related Work (1 page)\n3. Method (2 pages)\n4. Experiments (2.5 pages)\n5. Ablation \u002F Analysis (1 page)\n6. Conclusion (0.5 pages)\n```\n\n### Step 3: Section-by-Section Planning\n\nFor each section, specify:\n\n```markdown\n### §0 Abstract\n- **One-sentence problem**: [what gap this paper addresses]\n- **Approach**: [what we do, in one sentence]\n- **Key result**: [most compelling quantitative finding]\n- **Implication**: [why it matters]\n- **Estimated length**: 150-250 words\n- **Self-contained check**: can a reader understand this without the paper?\n\n### §1 Introduction\n- **Opening hook**: [1-2 sentences that motivate the problem]\n- **Gap**: [what's missing in prior work]\n- **Key questions**: [the research questions this paper answers]\n- **Contributions**: [numbered list, matching Claims-Evidence Matrix]\n- **Hero figure**: [describe what Figure 1 should show — MUST include clear comparison if applicable]\n- **Estimated length**: 1.5 pages\n- **Key citations**: [3-5 papers to cite here]\n\n### §2 Related Work\n- **Subtopics**: [2-4 categories of related work]\n- **Positioning**: [how this paper differs from each category]\n- **Minimum length**: 1 full page (at least 3-4 paragraphs with substantive synthesis)\n- **Must NOT be just a list** — synthesize, compare, and position\n\n### §3 Method \u002F Setup \u002F Preliminaries\n- **Notation**: [key symbols and their meanings]\n- **Problem formulation**: [formal setup]\n- **Method description**: [algorithm, model, or experimental design]\n- **Formal statements**: [theorems, propositions if applicable]\n- **Proof sketch locations**: [which key steps appear here vs. appendix]\n- **Estimated length**: 1.5-2 pages\n\n### §4 Experiments \u002F Main Results\n- **Figures planned**:\n  - Fig 1: [description, type: bar\u002Fline\u002Ftable\u002Farchitecture, WHAT COMPARISON it shows]\n  - Fig 2: [description]\n  - Table 1: [what it shows, which methods\u002Fbaselines compared]\n- **Data source**: [which JSON files \u002F experiment results]\n\n### §5 Conclusion\n- **Restatement**: [contributions rephrased, not copy-pasted from intro]\n- **Limitations**: [honest assessment — reviewers value this]\n- **Future work**: [1-2 concrete directions]\n- **Estimated length**: 0.5 pages\n```\n\n### Step 4: Figure Plan\n\nList every figure and table:\n\n```markdown\n## Figure Plan\n\n| ID | Type | Description | Data Source | Priority |\n|----|------|-------------|-------------|----------|\n| Fig 1 | Hero\u002FArchitecture | System overview + comparison | manual | HIGH |\n| Fig 2 | Line plot | Training curves comparison | figures\u002Fexp_A.json | HIGH |\n| Fig 3 | Bar chart | Ablation results | figures\u002Fablation.json | MEDIUM |\n| Table 1 | Comparison table | Main results vs. baselines | figures\u002Fmain_results.json | HIGH |\n| Table 2 | Theory comparison | Prior bounds vs. ours | manual | HIGH (theory papers) |\n```\n\n**CRITICAL for Figure 1 \u002F Hero Figure**: Describe in detail what the figure should contain, including:\n- Which methods are being compared\n- What the visual difference should demonstrate\n- Caption draft that clearly states the comparison\n\n### Step 5: Citation Scaffolding\n\nFor each section, list required citations:\n\n```markdown\n## Citation Plan\n- §1 Intro: [paper1], [paper2], [paper3] (problem motivation)\n- §2 Related: [paper4]-[paper10] (categorized by subtopic)\n- §3 Method: [paper11] (baseline), [paper12] (technique we build on)\n```\n\n**Citation rules** (from claude-scholar + Imbad0202\u002Facademic-research-skills):\n1. NEVER generate BibTeX from memory — always verify via search or existing .bib files\n2. Every citation must be verified: correct authors, year, venue\n3. Flag any citation you're unsure about with `[VERIFY]`\n4. Prefer published versions over arXiv preprints when available\n\n### Step 6: Cross-Review with REVIEWER_MODEL\n\nSend the complete outline to GPT-5.4 xhigh for feedback:\n\n```\nmcp__codex__codex:\n  model: gpt-5.4\n  config: {\"model_reasoning_effort\": \"xhigh\"}\n  prompt: |\n    Review this paper outline for a [VENUE] submission.\n    [full outline including Claims-Evidence Matrix]\n\n    Score 1-10 on:\n    1. Logical flow — does the story build naturally?\n    2. Claim-evidence alignment — every claim backed?\n    3. Missing experiments or analysis\n    4. Positioning relative to prior work\n    5. Page budget feasibility (MAX_PAGES = main body to Conclusion end, excluding refs\u002Fappendix)\n\n    For each weakness, suggest the MINIMUM fix.\n    Be specific and actionable — \"add X\" not \"consider more experiments\".\n```\n\nApply feedback before finalizing.\n\n### Step 7: Output\n\nSave the final outline to `PAPER_PLAN.md` in the project root:\n\n```markdown\n# Paper Plan\n\n**Title**: [working title]\n**Venue**: [target venue]\n**Type**: [empirical\u002Ftheory\u002Fmethod]\n**Date**: [today]\n**Page budget**: [MAX_PAGES] pages (main body to Conclusion end, excluding references & appendix)\n**Section count**: [N] (must match the number of section files that will be created)\n\n## Claims-Evidence Matrix\n[from Step 1]\n\n## Structure\n[from Step 2-3, section by section]\n\n## Figure Plan\n[from Step 4, with detailed hero figure description]\n\n## Citation Plan\n[from Step 5]\n\n## Reviewer Feedback\n[from Step 6, summarized]\n\n## Next Steps\n- [ ] \u002Fpaper-figure to generate all figures\n- [ ] \u002Fpaper-write to draft LaTeX\n- [ ] \u002Fpaper-compile to build PDF\n```\n\n## Key Rules\n\n- **Large file handling**: If the Write tool fails due to file size, immediately retry using Bash (`cat \u003C\u003C 'EOF' > file`) to write in chunks. Do NOT ask the user for permission — just do it silently.\n\n- **Do NOT generate author information** — leave author block as placeholder or anonymous\n- **Be honest about evidence gaps** — mark claims as \"needs experiment\" rather than overclaiming\n- **Page budget is hard** — if content exceeds MAX_PAGES, suggest what to move to appendix\n- **MAX_PAGES counts main body only** — from first page to end of Conclusion. References and appendix are NOT counted.\n- **Venue-specific norms** — all three venues (ICLR\u002FNeurIPS\u002FICML) use `natbib` (`\\citep`\u002F`\\citet`)\n- **Claims-Evidence Matrix is the backbone** — every claim must map to evidence, every experiment must support a claim\n- **Figures need detailed descriptions** — especially the hero figure, which must clearly specify comparisons and visual expectations\n- **Section count is flexible** — 5-8 sections depending on paper type. Don't force content into a rigid 5-section template.\n\n## Acknowledgements\n\nOutline methodology inspired by [Research-Paper-Writing-Skills](https:\u002F\u002Fgithub.com\u002FMaster-cai\u002FResearch-Paper-Writing-Skills) (claim-evidence mapping), [claude-scholar](https:\u002F\u002Fgithub.com\u002FGalaxy-Dawn\u002Fclaude-scholar) (citation verification), and [Imbad0202\u002Facademic-research-skills](https:\u002F\u002Fgithub.com\u002FImbad0202\u002Facademic-research-skills) (claim verification protocol).\n",{},"main","def8531e65114c0fca8fb8551c1871ee0eed705c","skills\u002Fwanshuiyin\u002FAuto-claude-code-research-in-sleep\u002Fpaper-plan",24,"https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub","https:\u002F\u002Fskillshub.wtf","https:\u002F\u002Fgithub.com\u002FComeOnOliver","AI & Blockchain Engineer\r\n","2026-06-24 22:07:19",{"source":7,"skillName":5,"commands":40,"prompt":50,"zipAvailable":51},[41,44,47],{"tool":42,"command":43},"npx","npx skills add ComeOnOliver\u002Fskillshub --skill paper-plan",{"tool":45,"command":46},"bunx","bunx skills add ComeOnOliver\u002Fskillshub --skill paper-plan",{"tool":48,"command":49},"pnpm","pnpm dlx skills add ComeOnOliver\u002Fskillshub --skill paper-plan","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fwanshuiyin\u002FAuto-claude-code-research-in-sleep\u002Fpaper-plan\u002FSKILL.md",true,[],[54,66,76,85,91,98],{"skillNo":55,"name":56,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":57,"categories":58,"keywords":62,"publishedAt":26},"01M6FKJN13GMG4KY57J3QD8GY5","model-hierarchy-skill","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzscole\u002Fmodel-hierarchy-skill\u002Fmodel-hierarchy-skill\u002FSKILL.md",[59],{"categoryNo":60,"categoryName":61,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000023","Skill与智能体开发",[63,64,65],"model","model hierarchy","model hierarchy skill",{"skillNo":67,"name":68,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":69,"categories":70,"keywords":72,"publishedAt":26},"01M6FKJJJ6GCAMR4PPPRD12M5X","zoom-mcp","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzoom\u002Fskills\u002Fzoom-mcp\u002FSKILL.md",[71],{"categoryNo":60,"categoryName":61,"taxonomyType":20,"taxonomyTypeDesc":21},[73,74,75],"MCP","zoom","zoom mcp",{"skillNo":77,"name":78,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":26},"01M6FKJE3KYT3KPEZMEEHYKA6K","zoom-apps-sdk","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzoom\u002Fskills\u002Fzoom-apps-sdk\u002FSKILL.md",[],[82,83,84,74],"sdk","zoom apps sdk","apps",{"skillNo":86,"name":87,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":26},"01M6FKJ9EDH9Q5VWDF8HAJRSP0","windows","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzoom\u002Fskills\u002Fwindows\u002FSKILL.md",[],[87],{"skillNo":92,"name":93,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":94,"categories":95,"keywords":97,"publishedAt":26},"01M6FKJ666BEEFC3AE9GJ72Q93","whiteboard","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzoom\u002Fskills\u002Fwhiteboard\u002FSKILL.md",[96],{"categoryNo":60,"categoryName":61,"taxonomyType":20,"taxonomyTypeDesc":21},[93],{"skillNo":99,"name":100,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":101,"categories":102,"keywords":103,"publishedAt":26},"01M6FKJ21WFCDGTJQ2Q91T9DVW","websockets","https:\u002F\u002Fgithub.com\u002FComeOnOliver\u002Fskillshub\u002Fblob\u002Fmain\u002Fskills\u002Fzoom\u002Fskills\u002Fwebsockets\u002FSKILL.md",[],[100]]