galaxy-dawn/claude-scholar
45 skills11.9k installs224k starsGitHub
Install
npx skills add https://github.com/galaxy-dawn/claude-scholarSkills in this repo
1Citation Verificationcitation-verification is a Claude Code skill from galaxy-dawn/claude-scholar that verifies academic citations and paper metadata before referencing them in research, documentation, or product features. It integrates Semantic Scholar's free graph API for paper lookup and search plus arXiv endpoints, with Python examples using the semanticscholar package. Semantic Scholar allows roughly 100 requests per 5 minutes without an API key and returns rich metadata across disciplines. Developers reach for citation-verification when bibliographies, README citations, or AI feature docs must confirm paper titles, authors, and IDs against authoritative sources rather than hallucinated references.1.1kinstalls2Paper Self Reviewpaper-self-review is a Claude Code agent skill from the galaxy-dawn/claude-scholar repository that guides agents to critique academic manuscripts before journal or conference submission. The skill evaluates writing clarity, logical structure, alignment between claims and cited evidence, gaps in related work coverage, and consistency of figures and tables against the narrative. Developers and researchers reach for paper-self-review when a draft is feature-complete but needs a structured pre-submission pass that mimics external peer review without waiting weeks for referee reports. Agents produce reviewer-style critique lists highlighting weak arguments, missing citations, overstated conclusions, and sections that confuse readers—helping authors revise proactively. paper-self-review complements grammar tools by focusing on scientific rigor and exposition quality rather than spelling alone. Research engineers, PhD students, and technical authors publishing ML systems or software engineering papers benefit when agents must systematically stress-test a manuscript's argumentation before upload to arXiv or a program committee portal.475installs3Writing Anti Aiwriting-anti-ai is a galaxy-dawn/claude-scholar agent skill (version 1.0.0, ~36 KB SKILL.md) that humanizes AI-generated prose using Wikipedia's Signs of AI writing guide as its pattern catalog. The skill flags and rewrites inflated symbolism, promotional language, vague attributions, superficial -ing analyses, AI vocabulary, negative parallelisms, and excessive conjunctive phrases, then adjusts rhythm, specificity, and voice across English and Chinese drafts. A five-rule workflow covers cutting filler phrases, breaking formulaic structures, varying sentence rhythm, trusting readers, and removing quotable slop, with scoring rubrics and before/after examples bundled in the skill file. Developers reach for writing-anti-ai when polishing AI-assisted blog posts, documentation, or landing-page copy before publication, or when reviewers flag robotic tone and detector hits. Install with npx skills add galaxy-dawn/claude-scholar --skill writing-anti-ai for Claude Code, Cursor, or Codex agents.455installs4Review Responsereview-response is an academic writing skill in galaxy-dawn/claude-scholar that automates the peer-review response workflow from comment analysis through professional rebuttal drafting. The four-stage pipeline parses and classifies reviewer comments into Major, Minor, Typo, and Misunderstanding categories, develops Accept, Defend, Clarify, or Experiment strategies per comment, writes structured rebuttal documents with evidence anchors, and runs tone management for professionalism and respect. Triggers include write rebuttal, respond to reviewers, draft review response, and analyze review comments. The skill ships reference files for classification criteria, response strategies, and rebuttal templates, and pairs with a rebuttal-writer agent plus a /rebuttal command for quick-start workflows. Install with npx skills add galaxy-dawn/claude-scholar --skill review-response. Developers and researchers reach for review-response when turning raw reviewer PDFs or comment blocks into point-by-point rebuttal letters before journal resubmission, ensuring every comment receives a response with explicit unresolved items marked rather than hidden.379installs5Ui Ux Pro Maxui-ux-pro-max is a frontend design skill from galaxy-dawn/claude-scholar for developers implementing interfaces that need professional hierarchy, accessibility, and visual polish. The skill steers agents through layout decisions, component spacing, contrast, focus order, and pattern selection so shipped UI reads clearly on first use. Developers reach for ui-ux-pro-max when a feature branch has functional code but weak hierarchy, inconsistent spacing, or accessibility gaps that block release quality. ui-ux-pro-max complements implementation work by naming concrete UI refinements instead of leaving design judgment entirely to the model.373installs6Results Analysisresults-analysis is a research interpretation skill from galaxy-dawn/claude-scholar that helps developers and ML researchers evaluate experiment outputs including metrics, ablation comparisons, and statistical significance before committing to full paper drafts. The skill supports go-no-go decisions on whether reported results substantiate stated hypotheses or require additional runs, different baselines, or revised claims. Researchers reach for results-analysis after training runs or benchmark sweeps complete and before investing time in LaTeX manuscripts or conference submissions. Source documentation is minimal, so treat it as a structured analysis workflow rather than a domain-specific stats library integration.360installs7Results ReportClaude Scholar skill for producing rigorous academic-style results reports from research inputs, organizing methods, outcomes, and limitations into documentation suitable for papers, theses, or stakeholder briefs.318installs8Skill Quality ReviewerEvaluates Claude Code skills against quality standards, flagging vague triggers, missing examples, and unsafe patterns so agent tooling ships consistent, reliable, and maintainable capability definitions.311installs9Zotero Obsidian Bridgezotero-obsidian-bridge from galaxy-dawn/claude-scholar connects Zotero reference management with Obsidian note-taking. It automates moving bibliographic data, attachments, and reading notes into a linked vault so researchers maintain one searchable, citation-aware knowledge graph.301installs10Ml Paper Writingml-paper-writing helps researchers convert experiment notes into polished ML manuscripts: venue-appropriate structure, rigorous methods prose, results narrative, citations, and reviewer-conscious limitations sections.300installs11Kaggle Learnerkaggle-learner from claude-scholar supports hands-on ML research via Kaggle: dataset exploration, feature engineering ideas, baseline models, cross-validation plans, and structured iteration for competition-style problems.287installs12Daily Paper Generatordaily-paper-generator automates scholarly literature monitoring by fetching recent papers for defined interests, summarizing methods and findings, and delivering a structured digest researchers can skim to stay current.271installs13Bug Detectivebug-detective guides Claude through disciplined debugging of scholarly software: reproduce failures, inspect logs and traces, isolate minimal cases, and propose verified fixes for notebooks, training scripts, and analysis pipelines.266installs14Skill Improverskill-improver from galaxy-dawn/claude-scholar upgrades Claude agent skills by diagnosing weak triggers, vague instructions, missing edge cases, and bloated SKILL.md files. It rewrites skill metadata, examples, and bundled resources so downstream agents invoke the right capability at the right time with fewer misfires and clearer guardrails.260installs15Obsidian Literature Workflowobsidian-literature-workflow in claude-scholar sets up Obsidian for scholarly reading: templated notes, linked concepts, reading pipelines, and synthesis artifacts that support literature reviews and long-form research writing.257installs16Latex Conference Template OrganizerHelps researchers organize and adapt LaTeX conference submission templates end to end. It structures class files, section sources, figures, tables, and bibliography assets to match venue formatting rules, reducing compile failures and manual reformatting so authors can focus on content while meeting strict page, margin, and metadata requirements.254installs17Architecture Designarchitecture-design from claude-scholar helps teams produce coherent system architecture documents covering components, data flows, integration points, and tradeoffs so validate-phase decisions are explicit before frontend or backend implementation.251installs18Publication Chart Skillpublication-chart-skill produces camera-ready scholarly figures from experiment metrics—learning curves, bar comparisons, ablation tables as plots—with consistent typography, color rules, captions, and export settings for ML venues.250installs19Uv Package ManagerGuides use of the uv Python package manager for creating projects, syncing dependencies, managing virtual environments, and running scripts with reproducible lockfiles in modern Python backends.250installs20Daily Codingdaily-coding in claude-scholar structures repeatable daily practice for developers: session planning, focused tasks, progress notes, and habit loops that sustain momentum during active implementation.248installs21Doc CoauthoringGuides Claude through structured scholarly coauthoring for long-form academic and research documents. It supports outlining, section-by-section drafting, revision passes, and consistency across citations, terminology, and argument flow so teams can move from rough notes to submission-ready manuscripts without losing author voice or document structure.247installs22Git Workflowgit-workflow in claude-scholar teaches production Git habits: branching models, commit discipline, pull request narratives, rebasing, and conflict handling so changes ship through review with fewer integration surprises.246installs23Webapp Testingwebapp-testing from galaxy-dawn/claude-scholar guides systematic testing of browser-based applications tied to academic or research tooling. It structures functional checks, edge cases, and regression passes so web UIs behave correctly before release, demo, or peer review.246installs24Code Review ExcellenceGuides Claude through excellent code review: read diffs holistically, prioritize blocking vs nit issues, verify tests and docs, flag security/perf risks, and produce clear, kind, actionable feedback authors can apply quickly.245installs25Web Design Reviewerweb-design-reviewer provides a ship-phase design audit for websites and web apps, flagging hierarchy, spacing, responsiveness, accessibility, and branding problems so teams fix UX defects before launch instead of after user complaints.245installs26Verification Loopverification-loop from galaxy-dawn/claude-scholar implements a scholar-oriented verify-and-correct cycle for research writing. It helps Claude repeatedly check claims, citations, and reasoning against sources, then apply fixes until the draft meets academic rigor standards.243installs27Planning With Filesplanning-with-files teaches agents to manage build execution through structured project files—plans, task lists, and status notes—so pm-style coordination stays durable, inspectable, and reusable across long implementation sessions.242installs28Post Acceptancepost-acceptance orchestrates the workflow after paper acceptance: camera-ready fixes, supplementary materials, reproducibility artifacts, repository releases, and distribution messaging so published work is complete and discoverable.242installs29Frontend Designfrontend-design from claude-scholar helps craft implementable UI and UX: information architecture, component patterns, visual hierarchy, responsive behavior, and accessibility considerations for web or app surfaces.241installs30Agent IdentifierHelps identify scholar-focused agents within galaxy-dawn/claude-scholar, classifying research assistants, corpora access patterns, and literature workflows so teams can credibly scope academic automation before deeper prototyping or integration.223installs31Defuddledefuddle simplifies intimidating academic material—dense abstracts, proofs, and domain jargon—into structured plain-language explanations with glossaries and prerequisite links so researchers grasp ideas faster.220installs32Mcp Integrationmcp integration is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted development.216installs33Hook Developmenthook development is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted development.213installs34Command Developmentcommand development is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted development.212installs35Plugin Structureplugin structure is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted development.211installs36Nature Polishingnature-polishing is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted development.190installs37Obsidian Project Kb Coreobsidian-project-kb-core scaffolds a linked Obsidian knowledge base for scholarly projects: folder layout, literature and experiment note templates, wikilinks, tags, and decision records that keep research context navigable.189installs38Obsidian Kb ArtifactsProduces Obsidian knowledge-base artifacts: structured notes, MOCs, literature summaries, and linked research pages with consistent frontmatter, wikilinks, and tags so personal or team vaults stay navigable and compounding.186installs39Obsidian Source Ingestionobsidian-source-ingestion from claude-scholar automates bringing external sources into Obsidian: PDF and web capture, bibliographic metadata, deduplication, tagging, and standardized note templates for a maintainable research vault.185installs40Nature WritingHelps create or rebuild Nature-style manuscript prose from author claims, results and notes without inventing evidence. A researcher uses it to write or restructure manuscript sections rather than just polish sentences.170installs41Expression SkillProvides a reusable communication workflow that prioritizes conclusion, evidence, risk and concrete action. A developer uses it for task reports, research discussion or high-signal responses.144installs42Nature ResponseConverts editor decision letters and reviewer comments into an auditable point-by-point response package mapping each concern to a manuscript change. A researcher uses it when responding to major or minor revision requests.144installs43Nature DataTurns a manuscript's supporting data into a Nature-ready data availability package including statement text, repository plans and dataset citations. A researcher uses it when preparing data-sharing sections for Nature-family submissions.143installs44Research IdeationRuns the research initiation workflow using the 5W1H framework, literature search, and gap analysis to produce a research question card and proposal. A researcher uses it when starting a project and needs to define and plan a defensible research question.35installs45Skill DevelopmentGuides authoring or repairing a Claude skill by locking its contract, writing lean frontmatter and workflow, and running integrity checks. A developer uses it to create a new skill or clean up one that is too long, too thin, or has stale references.24installs