菜单
Search arXiv for latest AI/ML preprints. Use when Claude needs bleeding-edge research not yet in peer-reviewed journals — newest methods, architectures, training techniques, benchmarks.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Web-grounded SOTA discovery using Perplexity Sonar models via OpenRouter. Use when academic APIs return foundational papers but you need to know what's CURRENT — what has superseded older methods, what practitioners actually use today, and what was released in the last weeks/months.
AI frontier plugin — gives Claude and agents access to latest scientific research, SOTA discovery, and architecture evaluation. Use when needing SOTA methods, research consensus, implementation guidance, or evaluating whether a codebase uses optimal patterns and architectures.
Search Hugging Face Papers for AI/ML research with AI-generated summaries, keywords, community upvotes, and trending papers. Use when Claude needs community-curated research, AI-generated paper summaries, or trending ML papers.
Search Semantic Scholar for papers with rich metadata — citation counts, AI-generated TLDRs, influence scores, citation graphs. Use for finding established research, tracing citation chains, understanding paper impact and field structure.
| name | arxiv-search |
| description | Search arXiv for latest AI/ML preprints. Use when Claude needs bleeding-edge research not yet in peer-reviewed journals — newest methods, architectures, training techniques, benchmarks. |
| keywords | ["arxiv","preprints","papers","research","machine-learning","deep-learning","AI","foundation-models"] |
Search arXiv preprints across AI/ML categories: cs.AI, cs.LG, cs.CL, cs.CV, stat.ML, cs.MA.
No API key required. Uses arXiv public API directly.
Find the script path, then execute:
SCRIPT=$(find ~/.claude/plugins -path "*/ai-frontier/skills/arxiv-search/scripts/search.mjs" 2>/dev/null | head -1)
node "$SCRIPT" "your query" [maxResults] [--sort=date|relevance] [--cats=cs.AI,cs.LG]
| Arg | Default | Description |
|---|---|---|
| query | (required) | Natural language search query |
| maxResults | 10 | Number of results (1-100) |
| --sort=date | relevance | Sort by submission date or relevance |
| --cats=cs.AI,cs.LG | all 6 defaults | Comma-separated arXiv categories |
--sort=date)Returns JSON with unified envelope:
results[].title — paper titleresults[].authors — author listresults[].abstract — first 500 chars of abstractresults[].pdf_url — direct PDF linkresults[].source_specific.arxiv_id — arXiv identifierresults[].source_specific.categories — arXiv categoriesresults[].source_specific.published — publication date# Latest papers on chain-of-thought reasoning
node "$SCRIPT" "chain-of-thought reasoning" 10 --sort=date
# Search specific category
node "$SCRIPT" "diffusion models" 5 --cats=cs.CV
# Broad AI safety search
node "$SCRIPT" "AI alignment safety" 20