[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M77VG3DTRSZ1C4N57KNFARD5":3,"skill-same:01M77VG3DTRSZ1C4N57KNFARD5":49},{"skillNo":4,"name":5,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":6,"repositoryNo":11,"starsCount":12,"licenseSpdxId":13,"licenseFileUrl":6,"sourceHtmlUrl":14,"categories":15,"keywords":16,"publishedAt":17,"linkOnly":18,"markdownContent":19,"markdownBody":19,"frontMatter":20,"refName":21,"commitSha":22,"skillDirectory":23,"forksCount":24,"repositoryHtmlUrl":25,"repositoryHomepageUrl":26,"ownerHtmlUrl":27,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":28,"install":29,"resources":42},"01M77VG3DTRSZ1C4N57KNFARD5","academic-paper-generation",null,"CorvinLabs\u002FCorvinOS","CorvinLabs","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F289506478?v=4","01M77VCHYED8C391HPBTNYP0HS","01M77VCHZ5K92DH3W6SPN8Y4P2",25,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FCorvinLabs\u002FCorvinOS\u002Fblob\u002Fmain\u002Fcorvin_operator\u002Fskills\u002Facademic-paper-generation\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"# Academic Paper Generation: From Concepts to Production-Ready PDF\n\n**Scope:** Complete workflow for generating publication-ready academic papers  \n**Audience:** Academic writing, arXiv submissions, conference papers  \n**Complexity:** High (multi-phase orchestration)  \n**Time Investment:** Significant initial learning, fast payoff on repeated use\n\n---\n\n## Overview\n\nThis skill provides a systematic, reusable workflow for transforming raw technical concepts, architectural decisions (ADRs), and research findings into a complete, polished academic paper suitable for arXiv, IEEE, ACM, or other venues.\n\nThe workflow enforces quality gates at each phase to catch common academic writing failures early:\n- Weak narrative arc (no clear story)\n- Thin content (placeholders instead of depth)\n- Broken references (placeholder citations without research)\n- Poor figure quality (artifacts, misalignment, missing context)\n- Structural inconsistency (section depth varies)\n\n---\n\n## The Five-Phase Workflow\n\n### Phase 1: Research & Reference Collection (Narrative Planning)\n\n**Goal:** Establish the paper's core claim and gather supporting references.\n\n**Steps:**\n1. **Read source material** (ADRs, concepts, existing docs) and extract the central insight—the one-sentence claim the paper argues.\n   - Good claim: \"Compliance mechanisms must be structural invariants, not configurable features.\"\n   - Bad claim: \"We implemented twelve compliance mechanisms.\"\n\n2. **Identify the problem space** via three historical examples (real incidents, not invented):\n   - What went wrong?\n   - Why did it happen? (root cause)\n   - What does it reveal about current practice?\n   \n3. **Dispatch reference research agents** (parallel work):\n   - Agent A: \"Find 15 academic papers on [topic + regulatory domain]\"\n   - Agent B: \"Find 5 industry incident postmortems (Equifax, Facebook, etc.)\"\n   - Agent C: \"Find 10 standards\u002Fregulations (GDPR, EU AI Act, HIPAA, PCI-DSS)\"\n   \n   Each agent returns curated list with URLs, citations, and 1-line summary of relevance.\n\n4. **Build a reference database** (CSV or JSON):\n   ```\n   category | author | year | title | url | relevance | sections_to_cite\n   ----------|--------|------|-------|-----|-----------|------------------\n   regulatory | EU | 2024 | EU AI Act | ... | Article 50 | intro, m1\n   incident | Rouse | 2017 | Equifax | ... | compliance failure | intro\n   academic | Emerson | 2008 | Model Checking | ... | formal methods | methodology\n   ```\n\n5. **Map references to paper sections** (pre-commitment):\n   - Introduction (3-5 foundational + 2-3 incident refs)\n   - Methodology (3-4 formal methods refs)\n   - Each mechanism (1-2 regulatory + 1-2 domain-specific refs)\n   - Related Work (5-7 competitive\u002Fcomplementary papers)\n\n**Deliverable:** A reference database and a one-paragraph problem statement that explains the paper's narrative arc.\n\n**Quality Gate:** Every section has ≥3 references allocated. No section is empty.\n\n---\n\n### Phase 2: Narrative Arc Design (Story Structure)\n\n**Goal:** Create a clear, compelling story that carries the reader from problem through solution to validation.\n\n**The IMRAD+C Structure (adapted for this domain):**\n\n1. **Introduction (40% of body text)**\n   - Hook (problem statement, 3 incidents showing pattern)\n   - Gap (current approaches fail because...)\n   - Thesis (our solution: structural invariants)\n   - Roadmap (in this paper we...)\n\n2. **Methodology (15% of body text)**\n   - Formal model (state machines, CTL*)\n   - Threat model (what could break compliance)\n   - Design principles (fail-closed, audit-first, etc.)\n\n3. **Results \u002F Mechanisms (30% of body text)**\n   - The twelve mechanisms, each:\n     - Regulatory requirement (cite the article)\n     - Implementation (how we enforce it)\n     - Why it's structural (can't be toggled away)\n     - Verification (how we prove it works)\n\n4. **Application \u002F Deployment (10% of body text)**\n   - CorvinOS as proof-of-concept\n   - Production stats (10M+ events, 50+ tenants, 100% audit clean)\n   - Performance overhead (measured, not claimed)\n\n5. **Conclusion (5% of body text)**\n   - Restatement of thesis\n   - Implications (compliance as architecture, not configuration)\n   - Future work (formal verification, distributed systems)\n\n**Narrative Arc Check:**\n```\nProblem:  \"Compliance keeps getting disabled under operational pressure\" \n          ✓ Concrete examples (Equifax, Facebook, SolarWinds)\n          ✓ Root cause explained (treated as configurable feature)\n\nSolution: \"Make compliance a structural invariant instead\"\n          ✓ Formal definition\n          ✓ Twelve orthogonal mechanisms\n          ✓ Can't be disabled without system restart\n\nProof:    \"CorvinOS shows this works in production\"\n          ✓ 12 months, 50+ tenants, zero incidents\n          ✓ Measurable audit overhead\n          ✓ Real deployment data\n```\n\n**Deliverable:** A one-page outline with section titles, estimated word count per section, and narrative flow check.\n\n**Quality Gate:** Reader can understand the paper's claim by reading only the section headings. Every section advances the narrative.\n\n---\n\n### Phase 3: Content Generation (Writing with Depth)\n\n**Goal:** Write sections with sufficient depth (3-5 well-developed paragraphs, not 1-2 sentences).\n\n**Heuristics for Content Density:**\n\n| Section | Length | Paragraphs | References | Formulas |\n|---------|--------|-----------|-----------|----------|\n| Abstract | 150-200 words | 1 | 0 | 0 |\n| Introduction | 800-1200 words | 4-6 | 5-8 | 1-2 |\n| Methodology | 600-1000 words | 3-5 | 3-5 | 3-5 |\n| Each Mechanism | 200-300 words | 2-3 | 1-2 | 0-1 |\n| Related Work | 400-600 words | 3-4 | 7-10 | 0 |\n| Conclusion | 300-400 words | 2-3 | 0-2 | 0 |\n\n**Writing Patterns:**\n\nFor each mechanism (M1-M12), follow this template:\n\n```\n**Regulatory Basis:** \"GDPR Article X requires...\"\n→ Cite the regulation, explain what it's trying to protect\n\n**Current Practice Fails:** \"Most systems implement this as...\"\n→ Show why typical implementations are configurable\u002Foptional\n\n**Structural Solution:** \"In CorvinOS, we enforce this by...\"\n→ Explain the architectural decision\n→ Show why it cannot be disabled\n\n**Verification:** \"Auditors can verify by...\"\n→ Concrete, measurable verification method\n→ Complexity (O(n), O(1), etc.)\n\n**Example Impact:** \"When [scenario], the system [response]\"\n→ Concrete scenario showing the invariant holding\n```\n\n**References Within Text:**\n\n- Cite regulations by article number: \"GDPR Article 30 requires...\"\n- Cite academic work early in sections: \"Prior work on compliance-as-code [Ref] assumes...\"\n- Cite incidents in introduction: \"Equifax (2017) disabled audit logging [Ref]...\"\n- Do NOT use placeholder citations like \\cite{ComplianceAsCode2019}; research the paper first.\n\n**Deliverable:** Complete draft of all sections with proper citations embedded.\n\n**Quality Gate:** \n- No \\cite{} placeholders—every citation is researched and valid\n- Minimum 2-3 paragraphs per mechanism\n- Every regulatory requirement is cited to the actual regulation\n- Abstract and conclusion are polished and final\n\n---\n\n### Phase 4: Figures & Diagrams (Quality Visual Communication)\n\n**Goal:** Create publication-quality diagrams that support the narrative.\n\n**Diagram Types & Creation Strategy:**\n\n| Diagram | Type | Tools | Format | Checklist |\n|---------|------|-------|--------|-----------|\n| Threat Model | Flow | Mermaid \u002F Graphviz | SVG → PDF (not PNG) | Labels clear, no overlaps |\n| Hash Chain Timeline | Sequence | TikZ \u002F draw.io | PDF | Time axis labeled, colors distinct |\n| Compliance Matrix | Table | TikZ \u002F Excel → TikZ | PDF | Green=compliant, labels aligned |\n| Multi-tenant Isolation | Architecture | Mermaid \u002F draw.io | PDF | Boundaries clear, isolation obvious |\n| Mechanism Overview | Infographic | Custom SVG | PDF | Hierarchical layout, legend |\n\n**Critical Rules for Figure Quality:**\n\n1. **SVG → PDF Pipeline (not PNG):**\n   - SVG source files stored in `figures\u002F` directory\n   - Use `inkscape --export-type=pdf` to convert (no rsvg-convert issues)\n   - Verify PDF in Acrobat; check for text rendering issues\n   - PNG export is only fallback if PDF fails\n\n2. **Caption & Labeling:**\n   - Caption format: \"Figure N: [One-line claim]. [Supporting detail]. [How it relates to narrative].\"\n   - Example: \"Figure 2: Boot-Time Tripwire Sequence. Audit chain is verified before plugins load, ensuring no instance runs with corrupted records. This structural enforcement prevents compliance drift over time.\"\n   - All labels in figures must be English, sans-serif font (Arial, Helvetica)\n   - No diagrams use red\u002Fgreen alone (colorblind-safe palette)\n\n3. **Placement & Integration:**\n   - Figures referenced in text before appearing: \"As shown in Figure 2...\"\n   - No floating figures; every figure has a cite\u002Freference in text\n   - Maximum one figure per subsection; typically 1-2 per major section\n   - Total figures: 3-5 for a short paper, 8-12 for a long paper\n\n4. **Anti-patterns to Avoid:**\n   - No screenshots of code (use verbatim blocks instead)\n   - No auto-generated diagrams without manual cleanup\n   - No figures with blurry text or pixel artifacts\n   - No diagrams that replicate content in the text (figures should add, not duplicate)\n\n**Deliverable:** 4-6 publication-quality diagrams (PDF format) with captions and references integrated into text.\n\n**Quality Gate:** \n- Every diagram is in PDF, not PNG\n- Every diagram has a caption explaining narrative relevance\n- Every diagram is referenced by text (not orphaned)\n- No text rendering issues; all labels are crisp\n\n---\n\n### Phase 5: LaTeX Assembly & Compilation (Final Production)\n\n**Goal:** Produce a final, polished PDF suitable for submission.\n\n**Template & Structure:**\n\n```latex\n\\documentclass[11pt,a4paper]{article}\n\\usepackage{arxiv}\n\\usepackage[utf8]{inputenc}\n\\usepackage{hyperref}\n\\usepackage{amsmath}\n\\usepackage{graphicx}\n\n\\title{Title}\n\\author{Author1 \\and Author2}\n\\date{Date}\n\n\\begin{document}\n\\maketitle\n\\begin{abstract}...\\end{abstract}\n\\section{Introduction}...\\end{section}\n\\section{Methodology}...\\end{section}\n\\section{Results}...\\end{section}\n\\section{Conclusion}...\\end{section}\n\\begin{thebibliography}{99}...\\end{thebibliography}\n\\end{document}\n```\n\n**BibTeX Best Practices:**\n\n```bibtex\n@article{AuthorYear,\n  author = \"First Last and Second Last\",\n  title = \"Full Title with Capitals\",\n  journal = \"Journal Name\",\n  year = \"2024\",\n  volume = \"10\",\n  pages = \"123--456\"\n}\n\n@misc{Regulation2024,\n  title = \"Regulation Title\",\n  author = \"Organization\",\n  year = \"2024\",\n  note = \"Official Journal Reference\"\n}\n\n@techreport{CISAReport,\n  author = \"CISA\",\n  title = \"Full Title\",\n  year = \"2020\",\n  note = \"URL or Report Number\"\n}\n```\n\n**Compilation Workflow:**\n\n```bash\n# Step 1: Check for undefined references\ntectonic paper.tex 2>&1 | grep -E \"undefined|Missing\"\n\n# Step 2: Verify figure paths\nfor fig in figures\u002F*.pdf; do\n  pdfinfo \"$fig\" > \u002Fdev\u002Fnull || echo \"✗ $fig is corrupted\"\ndone\n\n# Step 3: Full compilation\ntectonic paper.tex --keep-logs\n\n# Step 4: Verify output\npdfinfo paper.pdf | head -15\n```\n\n**Final Checks Before Submission:**\n\n- [ ] Abstract is 150-200 words, standalone (no references to sections)\n- [ ] All figures are in PDF, not PNG\n- [ ] No \\cite{} placeholders (every citation resolved)\n- [ ] Page count is reasonable (4-8 pages for short paper, 12-20 for full paper)\n- [ ] Margins are 1 inch (0.75-1 inch is acceptable)\n- [ ] All figures have captions and are referenced in text\n- [ ] Keywords are listed (typically 5-7)\n- [ ] No typos in author names or affiliations\n- [ ] References are sorted consistently (alphabetical or numeric)\n\n**Deliverable:** A final paper.pdf, production-ready and suitable for arXiv\u002Fjournal submission.\n\n**Quality Gate:**\n- PDF compiles without errors or warnings\n- All figures render correctly (pdfinfo validates)\n- Page count is reasonable\n- All references are resolved\n- Visual inspection passes (no typos, formatting is clean)\n\n---\n\n## Anti-Patterns to Avoid\n\n| Anti-Pattern | Why It Fails | Fix |\n|--------------|-------------|-----|\n| Placeholder citations (\\cite{ComplianceAsCode2019} without research) | Reader can't verify; looks unfinished | Research first, cite real papers with URLs |\n| One-sentence mechanism descriptions | Lacks depth; looks like a checklist, not analysis | 2-3 paragraphs per mechanism; regulatory basis + implementation + why structural |\n| PNG figures with artifacts | Blurry, unprofessional; looks auto-generated | Use SVG → PDF pipeline; verify in Acrobat |\n| No narrative arc (just lists facts) | Reader doesn't understand why they should care | IMRAD structure: problem → solution → proof → implications |\n| Figures without captions (self-explanatory claim) | Impossible to reference; violates journal guidelines | Every figure gets a caption explaining narrative relevance |\n| Section depth varies wildly (1 para, then 5 paras) | Looks inconsistent and unprofessional | Target density: 3-5 paras per major section |\n| Related work crammed into conclusion | Makes conclusion weak; related work hard to find | Dedicated Related Work section (one major section) |\n| No production data (only theory) | Claims unsupported; looks academic but not grounded | Include real deployment stats (page count, latency, audit events) |\n\n---\n\n## Implementation Checklist\n\nUse this checklist to track progress through all five phases:\n\n- [ ] **Phase 1: Research**\n  - [ ] Central claim identified and phrased\n  - [ ] 3+ historical examples (incidents) researched\n  - [ ] Reference database built (40+ sources)\n  - [ ] References mapped to sections\n\n- [ ] **Phase 2: Narrative**\n  - [ ] One-page outline written\n  - [ ] IMRAD structure mapped\n  - [ ] Narrative arc checked (problem → solution → proof)\n  - [ ] Section word counts estimated\n\n- [ ] **Phase 3: Content**\n  - [ ] All sections written (no placeholders)\n  - [ ] 3-5 paragraphs per major section\n  - [ ] All citations researched (no \\cite{} placeholders)\n  - [ ] Examples and scenarios concrete (not hypothetical)\n\n- [ ] **Phase 4: Figures**\n  - [ ] 4-6 diagrams created (SVG source files saved)\n  - [ ] All diagrams converted to PDF\n  - [ ] Captions written and embedded\n  - [ ] Figures referenced in text\n\n- [ ] **Phase 5: Assembly**\n  - [ ] LaTeX template created\n  - [ ] BibTeX references formatted correctly\n  - [ ] Paper compiles without errors\n  - [ ] Final checks passed\n\n---\n\n## Automation Opportunities\n\nThis workflow can be partially automated:\n\n1. **Phase 1 (Research):** Dispatch three parallel research agents; aggregate results into reference DB\n2. **Phase 2 (Narrative):** Use template-driven outline generation; validate structure via checklist\n3. **Phase 3 (Content):** Use generation agents with narrative-aware prompts; verify density heuristics\n4. **Phase 4 (Figures):** Generate diagram specs from mechanisms; batch-convert SVG → PDF via Inkscape\n5. **Phase 5 (Assembly):** Template-driven LaTeX generation; automated compilation and validation\n\n**Next Skill:** \"Academic Paper Generator\" (orchestrates all five phases via multi-agent workflow)\n\n---\n\n## Resources & References\n\n- **ArXiv Best Practices:** https:\u002F\u002Farxiv.org\u002Fhelp\u002Fsubmit\n- **Henning, A. (2020). \"How to Write a Good Research Paper.\"** IEEE Computer Society\n- **Academic Writing Guides:** Purdue OWL (online writing lab)\n- **Diagram Tools:** Mermaid, Graphviz, TikZ, draw.io\n- **Compilation:** Tectonic (robust TeX engine)\n\n---\n\n## Key Takeaway\n\n**Compliance-as-Architecture, not Configuration** applies to papers too: structure the narrative arc, depth, figures, and citations as load-bearing elements of the paper, not as optional flourishes added at the end. A paper with a clear story, sufficient depth, and strong visuals is far more likely to be read, cited, and accepted than one where these elements are afterthoughts.\n",{},"main","3f28cfd9b82326ff325e645ff041589c3220d813","corvin_operator\u002Fskills\u002Facademic-paper-generation",4,"https:\u002F\u002Fgithub.com\u002FCorvinLabs\u002FCorvinOS","https:\u002F\u002Fcorvin-labs.com, https:\u002F\u002Fwww.youtube.com\u002F@Corvin-Labs","https:\u002F\u002Fgithub.com\u002FCorvinLabs","2026-11-14 23:30:13",{"source":7,"skillName":5,"commands":30,"prompt":40,"zipAvailable":41},[31,34,37],{"tool":32,"command":33},"npx","npx skills add CorvinLabs\u002FCorvinOS --skill academic-paper-generation",{"tool":35,"command":36},"bunx","bunx skills add CorvinLabs\u002FCorvinOS --skill academic-paper-generation",{"tool":38,"command":39},"pnpm","pnpm dlx skills add CorvinLabs\u002FCorvinOS --skill academic-paper-generation","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FCorvinLabs\u002FCorvinOS\u002Fblob\u002Fmain\u002Fcorvin_operator\u002Fskills\u002Facademic-paper-generation\u002FSKILL.md",true,[43],{"path":44,"resourceType":45,"resourceTypeDesc":46,"sizeBytes":47,"sourceHtmlUrl":48},"GETTING_STARTED.md","REFERENCE","参考文档",10744,"https:\u002F\u002Fgithub.com\u002FCorvinLabs\u002FCorvinOS\u002Fblob\u002Fmain\u002Fcorvin_operator\u002Fskills\u002Facademic-paper-generation\u002FGETTING_STARTED.md",[50,57,63,70,76,83],{"skillNo":51,"name":52,"nameZh":6,"description":53,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":6,"repositoryNo":11,"starsCount":12,"licenseSpdxId":13,"licenseFileUrl":6,"sourceHtmlUrl":54,"categories":55,"keywords":56,"publishedAt":17},"01M77VGAEV6EQA3C62F78EDHQ3","voice","Voice mode for Claude Code — automatically reads assistant replies aloud via TTS, with German\u002FEnglish auto-detection and optional summarization for long responses. Trigger when the user mentions voice, vorlesen, sprachausgabe, TTS, \"speak\", \"say it out loud\", or asks about \u002Fvoice-* commands.","https:\u002F\u002Fgithub.com\u002FCorvinLabs\u002FCorvinOS\u002Fblob\u002Fmain\u002Fcorvin_operator\u002Fvoice\u002Fskills\u002Fvoice\u002FSKILL.md",[],[],{"skillNo":58,"name":59,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":6,"repositoryNo":11,"starsCount":12,"licenseSpdxId":13,"licenseFileUrl":6,"sourceHtmlUrl":60,"categories":61,"keywords":62,"publishedAt":17},"01M77VG6N69YBC0CR0TSMGXHTV","paper-generator-orchestrator","https:\u002F\u002Fgithub.com\u002FCorvinLabs\u002FCorvinOS\u002Fblob\u002Fmain\u002Fcorvin_operator\u002Fskills\u002Fpaper-generator-orchestrator\u002FSKILL.md",[],[],{"skillNo":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":6,"repositoryNo":11,"starsCount":12,"licenseSpdxId":13,"licenseFileUrl":6,"sourceHtmlUrl":67,"categories":68,"keywords":69,"publishedAt":17},"01M77VFXA58C38HG5K5J4Q67MF","forge","Forge schema-bound, sandboxed tools at runtime. Paired with the cowork `forge` persona for chat-driven use in the voice-skill bridge. Use when you would otherwise write the same Bash snippet ≥3× with different parameters, when you need precise numerical results over a dataset, or when the user wants reproducible\u002Fdeterministic output written to a file.","https:\u002F\u002Fgithub.com\u002FCorvinLabs\u002FCorvinOS\u002Fblob\u002Fmain\u002Fcorvin_operator\u002Fforge\u002FSKILL.md",[],[],{"skillNo":71,"name":72,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":6,"repositoryNo":11,"starsCount":12,"licenseSpdxId":13,"licenseFileUrl":6,"sourceHtmlUrl":73,"categories":74,"keywords":75,"publishedAt":17},"01M77VFTJ8KS80R6PVPTBH4CWV","cowork","https:\u002F\u002Fgithub.com\u002FCorvinLabs\u002FCorvinOS\u002Fblob\u002Fmain\u002Fcorvin_operator\u002Fcowork\u002Fskills\u002Fcowork\u002FSKILL.md",[],[],{"skillNo":77,"name":78,"nameZh":6,"description":79,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":6,"repositoryNo":11,"starsCount":12,"licenseSpdxId":13,"licenseFileUrl":6,"sourceHtmlUrl":80,"categories":81,"keywords":82,"publishedAt":17},"01M77VFQTC6WR708YMVHAKQWWN","quality-layer-control","Toggle quality disciplines (ADR Gate, docs-as-definition-of-done, etc.) on or off globally","https:\u002F\u002Fgithub.com\u002FCorvinLabs\u002FCorvinOS\u002Fblob\u002Fmain\u002Fcorvin_operator\u002Fbundle\u002Fskills\u002Fquality-layer-control\u002FSKILL.md",[],[],{"skillNo":84,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":6,"repositoryNo":11,"starsCount":12,"licenseSpdxId":13,"licenseFileUrl":6,"sourceHtmlUrl":87,"categories":88,"keywords":89,"publishedAt":17},"01M77VFHGGPAQY8ES364WB7GDZ","using-ldd","Use at the start of any conversation, and whenever the user mentions LDD, loss-driven development, loss, gradient, SGD on code, drift, refinement loop, outer loop, inner loop, method evolution, or says \"LDD:\" \u002F \"apply LDD\". Establishes how to dispatch the ten LDD skills and when each fires. Must be the first skill considered in any coding task where LDD is available.","https:\u002F\u002Fgithub.com\u002FCorvinLabs\u002FCorvinOS\u002Fblob\u002Fmain\u002Fcorvin_operator\u002Fbundle\u002Fskills\u002Fldd\u002Fusing-ldd\u002FSKILL.md",[],[]]