All authors
Claude Skills by Soceremite
github.com/Soceremite197 skills0 installs33 views
- Research RefineTurn a vague research direction into a problem-anchored, elegant, frontier-aware, implementation-oriented method plan via iterative Gemini review. Use when the user says \"refine my approach\", \"帮我细化方案\", \"decompose this problem\", \"打磨idea\", \"refine research plan\", \"细化研究方案\", or wants a concrete research method that stays simple, focused, and top-venue ready instead of a vague or overbuilt idea.Votes: 0GitHub stars: 3
- Research ReviewGet a deep critical review of research from Gemini via gemini-review MCP. Use when user says \"review my research\", \"help me review\", \"get external review\", or wants critical feedback on research ideas, papers, or experimental results.Votes: 0GitHub stars: 3
- Ablation PlannerPlan and implement controlled or retuned ablations for submission after result-to-claim reports claim_supported=yes or partial. Use to isolate contributions and test alternatives.Votes: 0GitHub stars: 3
- AlphaxivExplain one paper from an arXiv/AlphaXiv URL or arXiv ID, using AlphaXiv with source fallback. Use for single-paper understanding and summaries.Votes: 0GitHub stars: 3
- Analyze ResultsAnalyse experiment metrics and sample-level behavior, compare methods fairly, and diagnose successes, regressions and failures. Use for result interpretation, error analysis or case studies.Votes: 0GitHub stars: 3
- ArxivSearch arXiv preprints, retrieve metadata, download PDFs, and summarize papers. Use when arXiv search or saving papers to a local library is requested.Votes: 0GitHub stars: 3
- Auto Paper Improvement LoopImprove a generated paper through bounded review, revision, and recompilation rounds. Use for iterative paper polishing after an initial draft exists.Votes: 0GitHub stars: 3
- Auto Review Loop LlmRun iterative research review and revision with an OpenAI-compatible LLM API. Use when that reviewer backend is requested; requires llm-chat or API setup.Votes: 0GitHub stars: 3
- Auto Review Loop MinimaxRun iterative research review and revision through the MiniMax API. Use when MiniMax is the requested external reviewer; requires configured credentials.Votes: 0GitHub stars: 3
- Auto Review LoopRun repeated research review, implement fixes, and re-review until the assessment or round limit stops the loop. Use for autonomous iterative improvement.Votes: 0GitHub stars: 3
- Citation AuditVerify bibliography entries and citation context against real sources. Use to detect fabricated references, attribution errors, or unsupported paper citations.Votes: 0GitHub stars: 3
- Claims DraftingDraft patent claims from an invention disclosure. Use for independent and dependent claims, claim scope, and formal claim language in a patent application.Votes: 0GitHub stars: 3
- Comm Lit ReviewReview communications research literature using the local knowledge base first. Use for related work in wireless, networking, satellite, MAC/PHY, and transport.Votes: 0GitHub stars: 3
- DeepxivSearch and progressively read papers with DeepXiv, including section-level access and trending papers. Use when DeepXiv-backed retrieval is requested.Votes: 0GitHub stars: 3
- Dse LoopRun bounded design-space exploration for computer architecture or EDA: evaluate, analyze, tune parameters, and repeat toward an objective or timeout.Votes: 0GitHub stars: 3
- Embodiment DescriptionWrite detailed patent embodiments explaining how to practice an invention. Use to turn technical implementations into specification-ready descriptions.Votes: 0GitHub stars: 3
- Environment MaintenanceInspect or maintain research environments through independent candidates, validation receipts, frozen revisions and consumer/lease/pin protection. Use for dependency changes, environment reuse and retirement planning.Votes: 0GitHub stars: 3
- Exa SearchSearch the web and extract content with Exa, including similar pages and broad sources beyond academic indexes. Use when Exa-backed retrieval is requested.Votes: 0GitHub stars: 3
- Experiment AuditAssess experiment integrity and comparative fairness from supplied records, or obtain an independent execution audit. Use for evaluation design checks, completed-run integrity review and evidence-based audit findings.Votes: 0GitHub stars: 3
- Experiment BridgeImplement an existing experiment plan, sanity-check code, deploy where available, and collect initial results. Use to move from an approved plan to execution.Votes: 0GitHub stars: 3
- Experiment PlanDesign fair comparative experiments for a fixed research proposal: controls, data and training budgets, tuning, seeds, evaluation, ablations and run order. Use for an experiment roadmap or comparison design.Votes: 0GitHub stars: 3
- Experiment QueueQueue multi-seed or multi-config ML jobs over SSH with OOM retry and safe wave transitions. Use for batch orchestration beyond a single experiment launch.Votes: 0GitHub stars: 3
- Feishu NotifySend requested Feishu/Lark notifications or authorized workflow status updates. Use with configured notification credentials and an explicit delivery scope.Votes: 0GitHub stars: 3
- Figure DescriptionDescribe supplied patent drawings with formal figure references and numbered components. Use to prepare drawing descriptions for a patent specification.Votes: 0GitHub stars: 3
- Figure SpecRender structured FigureSpec JSON into deterministic, editable SVG architecture or workflow diagrams. Use for precise publication figures and vector output.Votes: 0GitHub stars: 3
- Formula DerivationDevelop a coherent research derivation: choose mathematical objects, organize assumptions, and connect equations. Use before a theorem or proof target is fixed.Votes: 0GitHub stars: 3
- Gemini SearchDiscover research literature through Gemini. Use for explicitly requested Gemini paper search or broad AI-assisted discovery; requires a configured Gemini backend.Votes: 0GitHub stars: 3
- Grant ProposalDraft research funding proposals from ideas and literature in funder-specific or generic formats, including NSFC, NSF, ERC, and KAKENHI. Use for grant applications.Votes: 0GitHub stars: 3
- Idea CreatorGenerate and assess concrete research contributions within an established problem. Use for method candidates; scientific-brainstorming frames a vague direction and idea-discovery runs the complete workflow.Votes: 0GitHub stars: 3
- Idea Discovery RobotRun robotics or embodied-AI idea discovery with benchmark-grounded, simulation-first validation. Use for robot learning, manipulation, navigation, or locomotion.Votes: 0GitHub stars: 3
- Idea DiscoveryRun the complete idea-discovery workflow: literature, ideation, novelty checks, review, and pilots. Use when the full direction-exploration process is requested.Votes: 0GitHub stars: 3
- Integrity ForensicsRun SHA-pinned deterministic numeric/rules integrity forensics and record policy obligations. Codex mode can flag issues but cannot certify CLEAN or full semantic coverage.Votes: 0GitHub stars: 3
- Interview CheatsheetWrite a detailed Chinese ML/LLM interview guide with derivations, PyTorch examples, comparisons, and graded practice questions. Use for topic-specific interview prep.Votes: 0GitHub stars: 3
- Invention StructuringTurn a rough invention into a structured invention disclosure: technical problem, solution, distinguishing features, and effects. Use before patent drafting.Votes: 0GitHub stars: 3
- Jurisdiction FormatCompile a patent application into CN, US, or EP filing format. Use for jurisdiction-specific organization and formatting after application content exists.Votes: 0GitHub stars: 3
- Kill ArgumentRun adversarial rejection and defense reviews in separate fresh contexts. Use to surface unresolved weaknesses before submission or prepare for hostile reviewers.Votes: 0GitHub stars: 3
- Mermaid DiagramCreate and verify Mermaid diagrams from requirements, saving editable .mmd and .md files. Use for flowcharts, sequences, classes, ER diagrams, or timelines.Votes: 0GitHub stars: 3
- Meta ApplyLand approved meta-optimize patches to the skill corpus through its fresh landing-review and human-approval gate. Use when applying staged toolkit optimizations.Votes: 0GitHub stars: 3
- Meta OptimizeAnalyze toolkit usage logs and propose changes to skill instructions, reviewer prompts, or workflow defaults. Use for evidence-based harness optimization.Votes: 0GitHub stars: 3
- Monitor ExperimentCheck a running experiment once, inspect progress, and collect available outputs. Use for status or result queries; use training-check for continued monitoring.Votes: 0GitHub stars: 3
- Novelty CheckCheck a research idea against recent prior work and identify closest overlaps. Use to assess novelty before implementation; requires literature evidence.Votes: 0GitHub stars: 3
- OpenalexSearch OpenAlex for papers, citation links, institutions, and funding metadata. Use when open academic metadata or the OpenAlex citation graph is requested.Votes: 0GitHub stars: 3
- Overleaf SyncSynchronize a local paper directory with Overleaf in either direction. Use for explicitly requested pull, push, or project linking; requires Overleaf setup.Votes: 0GitHub stars: 3
- Paper Claim AuditCheck paper numbers, comparisons, and scope claims against raw result files with a fresh reviewer. Use for paper-to-evidence fidelity before submission.Votes: 0GitHub stars: 3
- Paper CompileCompile a LaTeX paper to PDF, repair compilation errors, and inspect the output. Use for an existing manuscript that needs a verified build.Votes: 0GitHub stars: 3
- Paper FigureCreate publication-quality data plots, tables and qualitative case panels from real experiment outputs. Use for paper result figures, visual comparisons or statistical displays.Votes: 0GitHub stars: 3
- Paper Illustration Image2Create paper illustrations through a local Codex app-server image bridge. Experimental alternative for Claude Code; requires its renderer setup.Votes: 0GitHub stars: 3
- Paper IllustrationCreate paper illustrations with Gemini image generation and iterative review. Use for AI-rendered method visuals; requires configured Gemini image tooling.Votes: 0GitHub stars: 3
- Paper PlanBuild a structured paper outline from results and review conclusions. Use to plan the narrative, sections, evidence, and figures before drafting.Votes: 0GitHub stars: 3
- Paper Poster HtmlBuild an academic poster as HTML/CSS with real paper figures, measured layout checks, and print-ready PDF export. Use for conference poster creation or redesign.Votes: 0GitHub stars: 3