[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-82161":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":13,"subscribersCount":13,"size":13,"stars1d":16,"stars7d":17,"stars30d":18,"stars90d":13,"forks30d":13,"starsTrendScore":19,"compositeScore":13,"rankGlobal":10,"rankLanguage":10,"license":20,"archived":21,"fork":21,"defaultBranch":22,"hasWiki":23,"hasPages":21,"topics":24,"createdAt":10,"pushedAt":10,"updatedAt":28,"readmeContent":29,"aiSummary":30,"trendingCount":13,"starSnapshotCount":13,"syncStatus":31,"lastSyncTime":32,"discoverSource":33},82161,"CiteCheck","color4-alt\u002FCiteCheck","color4-alt","Check academic paper citations for format, queryability, thematic relevance, and semantic accuracy.","",null,"Python",46,0,30,1,8,15,16,24,"MIT License",false,"main",true,[25,26,27],"academic-writing","agent-skills","citation-management","2026-06-12 02:04:23","\u003Cp align=\"center\">\n  \u003Cimg src=\"assets\u002Fbanner.jpg\" alt=\"CiteCheck Banner\" width=\"100%\">\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Ca href=\"README.md\">English\u003C\u002Fa> | \u003Ca href=\"README.zh.md\">中文\u003C\u002Fa>\n\u003C\u002Fp>\n\n# CiteCheck — Cross-Agent Citation Verification Skill\n\n**A portable agent skill + standalone CLI for verifying academic paper citations.**\n\nExtract references from LaTeX or PDF, validate formatting, verify existence via Crossref \u002F Semantic Scholar \u002F OpenAlex \u002F PubMed \u002F arXiv \u002F dblp \u002F Google Scholar \u002F WebSearch, and score thematic \u002F semantic relevance using cited paper abstracts — all without requiring an external LLM API key when used as a skill.\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fagents-Claude%20%7C%20Codex%20%7C%20OpenClaw%20%7C%20Hermes%20%7C%20Gemini%20%7C%20Cursor-blue\" alt=\"Supported Agents\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fskill%20standard-agentskills.io-green\" alt=\"Agent Skills Standard\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FPyPI-CiteCheck-orange?logo=pypi\" alt=\"PyPI\">\n\u003C\u002Fp>\n\n---\n\n## ✨ What is CiteCheck?\n\nCiteCheck is primarily a **cross-agent skill** that helps AI coding assistants verify citations in academic papers. It is designed to work across Claude Code, Codex, OpenClaw, Hermes, Gemini CLI, Cursor, and more — following the [agentskills.io](https:\u002F\u002Fagentskills.io) open standard.\n\nIt is also available as a **standalone Python CLI** for users who prefer running it directly from the terminal.\n\n> **Key Design Principle**: When used as a skill, thematic and semantic matching are performed directly by the host agent's own reasoning. No OpenAI API key is required. The CLI handles structured tasks (parsing, format checks, API queries) while the agent handles interpretive tasks (relevance scoring, claim verification).\n\n---\n\n## 🚀 Two Ways to Use\n\n### Mode 1: Agent Skill (Recommended)\n\nInstall CiteCheck as a skill for your coding agent. The agent will automatically discover and invoke it when you ask to check citations.\n\n**Step 1 — Install the skill**\n\n> 🟢 **Easiest way — just ask your agent:**\n>\n> ```\n> Install this skill for me: https:\u002F\u002Fgithub.com\u002Fcolor4-alt\u002FCiteCheck\n> ```\n>\n> Your agent will clone the repo into the correct skill directory automatically.\n\nIf you prefer to install manually:\n\n| Agent | Install Path |\n|-------|-------------|\n| **Claude Code** | `~\u002F.claude\u002Fskills\u002Fcitecheck` |\n| **Codex CLI** | `~\u002F.codex\u002Fskills\u002Fcitecheck` |\n| **OpenClaw** | `~\u002F.openclaw\u002Fskills\u002Fcitecheck` |\n| **Hermes** | `~\u002F.hermes\u002Fskills\u002Fcitecheck` |\n| **Gemini CLI** | `~\u002F.gemini\u002Fskills\u002Fcitecheck` |\n| **Cursor** | `.cursor\u002Frules\u002Fcitecheck.mdc` (copy `skills\u002Fcitecheck\u002FSKILL.md`) |\n| **GitHub Copilot** | Append `AGENTS.md` to `.github\u002Fcopilot-instructions.md` |\n\n**Step 2 — Invoke**\n\nUse natural language or a slash command with a file reference:\n\n```\n\u002Fcitation-verification @main.tex\n\u002Fcitation-verification @paper.pdf\n\u002Fcitation-verification @path\u002Fto\u002Flatex_project\u002F\n```\n\nOr simply tell your agent:\n\n```\nCheck the citations in this paper.\nVerify the references in my LaTeX project.\nAre these citations accurate and relevant?\n```\n\nThe agent will:\n1. Call `citecheck` CLI to parse the paper and check formatting\n2. Query Crossref → Semantic Scholar → OpenAlex → PubMed → arXiv → dblp → Google Scholar → WebSearch to verify paper existence\n3. Use its own reasoning to evaluate thematic relevance and semantic accuracy\n4. Present a structured Markdown report\n\n> **No API key needed.** The agent handles steps 3–4 with its built-in LLM capabilities.\n\n---\n\n### Mode 2: Standalone CLI\n\nFor users who prefer the command line or need to integrate into CI pipelines.\n\n**Step 1 — Install the Python package**\n\n```bash\npip install citecheck-cli\n```\n\nFor PDF support:\n\n```bash\npip install citecheck-cli[pdf]\n```\n\nOr install from source:\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Fcolor4-alt\u002FCiteCheck.git\ncd CiteCheck\npip install -e \".[pdf,dev]\"\n```\n\n**Step 2 — Run**\n\n```bash\n# Check a LaTeX project (preferred)\ncitecheck path\u002Fto\u002Flatex_project\u002F\n\n# Check a single .tex file\ncitecheck main.tex\n\n# Check a PDF (fallback)\ncitecheck paper.pdf -o report.md\n\n# Skip online verification (offline mode)\ncitecheck main.tex --skip-verification\n\n# Use external LLM for matching (requires --api-key)\ncitecheck main.tex --api-key $OPENAI_API_KEY\n```\n\n**CLI Options**\n\n```\ncitecheck [-h] [-o OUTPUT] [--skip-verification] [--skip-semantic] [--api-key API_KEY] [-v] input\n\npositional arguments:\n  input                 Path to paper (PDF, .tex, or directory with .tex + .bib)\n\noptions:\n  -o OUTPUT             Output report path (default: citation_check_report.md)\n  --skip-verification   Skip all online verification (Crossref \u002F Semantic Scholar \u002F OpenAlex \u002F PubMed \u002F arXiv \u002F dblp \u002F Google Scholar \u002F WebSearch)\n  --skip-semantic       Skip semantic matching\n  --api-key API_KEY     Optional OpenAI key for LLM matching (falls back to heuristics)\n  -v, --verbose         Verbose output\n```\n\n---\n\n## 📊 Workflow\n\n```\nInput (LaTeX \u002F PDF)\n    │\n    ▼\n┌─────────────────┐\n│ 1. Parse Paper  │  ← Extract refs, citations, body text\n└────────┬────────┘\n         │\n    ┌────┴────┐\n    ▼         ▼\n┌────────┐ ┌─────────────┐\n│ LaTeX  │ │ PDF Fallback │\n│(.bib)  │ │ (PyMuPDF)    │\n└────────┘ └─────────────┘\n         │\n         ▼\n┌─────────────────┐\n│ 2. Format Check │  ← Validate BibTeX fields, types, venues\n└────────┬────────┘\n         │\n         ▼\n┌─────────────────────┐\n│ 3. Queryability     │  ← Crossref → Semantic Scholar → OpenAlex → PubMed → arXiv → dblp → Google Scholar → WebSearch\n└────────┬────────────┘\n         │\n         ▼\n┌─────────────────────┐\n│ 4. Thematic Match   │  ← Skill: agent reasoning | CLI: heuristic\u002FLLM\n└────────┬────────────┘\n         │\n         ▼\n┌─────────────────────┐\n│ 5. Semantic Match   │  ← Skill: agent reasoning | CLI: heuristic\u002FLLM\n└────────┬────────────┘\n         │\n         ▼\n┌─────────────────────┐\n│ 6. Generate Report  │  ← Markdown report with findings\n└─────────────────────┘\n```\n\n---\n\n## 📋 Report Output\n\nCiteCheck generates a Markdown report containing:\n\n- **Summary**: total references, format issues, verified count, average scores\n- **Detailed table**: per-reference status for format \u002F queryable \u002F thematic \u002F semantic\n- **Format issues**: specific problems (missing author, wrong entry type, suspicious year, preprint source, etc.)\n- **Queryability results**: verification status from Crossref \u002F Semantic Scholar \u002F OpenAlex \u002F PubMed \u002F arXiv \u002F dblp \u002F Google Scholar \u002F WebSearch\n- **Abstract-aware semantic scoring**: semantic match uses the cited paper's abstract when available\n- **Uncited references**: entries in `.bib` never referenced by `\\cite{}` in the text\n\nSee [`examples\u002Fexample_report.md`](examples\u002Fexample_report.md) for a full sample.\n\n---\n\n## 🏗️ Project Structure\n\n```\nCiteCheck\u002F\n├── skills\u002Fcitecheck\u002FSKILL.md      ← Agent skill entry (cross-platform)\n├── .claude-plugin\u002Fplugin.json     ← Claude Code marketplace metadata\n├── .codex-plugin\u002Fplugin.json      ← Codex CLI marketplace metadata\n├── CLAUDE.md                      ← Project context for Claude Code\n├── AGENTS.md                      ← Project context for Codex \u002F generic agents\n├── GEMINI.md                      ← Project context for Gemini CLI\n├── src\u002Fcitecheck\u002F                 ← Python CLI source\n│   ├── cli.py\n│   ├── parser.py\n│   ├── bibtex_parser.py\n│   ├── pdf_parser.py\n│   ├── verifier.py\n│   ├── matcher.py\n│   ├── models.py\n│   └── reporter.py\n├── references\u002F                    ← Skill reference docs\n│   ├── format-check-rules.md\n│   ├── api-reference.md\n│   ├── thematic-scoring-prompt.md\n│   └── semantic-matching-prompt.md\n├── tests\u002F\n├── examples\u002F\n└── README.md \u002F README.zh.md\n```\n\n---\n\n## 🛠️ Development\n\n```bash\n# Clone\ngit clone https:\u002F\u002Fgithub.com\u002Fcolor4-alt\u002FCiteCheck.git\ncd CiteCheck\n\n# Install in editable mode with dev dependencies\npip install -e \".[dev]\"\n\n# Run tests\npytest tests\u002F -v\n\n# Format & lint\nblack src\u002F tests\u002F\nruff check src\u002F tests\u002F\n```\n\n---\n\n## 🤝 Contributing\n\n- Any change to `skills\u002Fcitecheck\u002FSKILL.md` must remain **agent-agnostic** (no brand-specific language)\n- Skill content should work across Claude Code, Codex, OpenClaw, Hermes, and Gemini CLI\n- When adding CLI features, update both `src\u002Fcitecheck\u002Fcli.py` and the README\n\n---\n\n## 📝 Changelog\n\n### 0.1.1 (2026-05-28)\n\n**Fixed**\n- PDF Parser: Fixed year extraction for arXiv references — arXiv IDs like `arXiv:2004.05150` were incorrectly parsed as the publication year. Now prefers year at end of citation and skips arXiv ID patterns.\n- Verifier (Crossref): Added title similarity scoring, author overlap check, and false-match rejection. Crossref now evaluates all 3 candidates and rejects matches with similarity \u003C 0.2.\n\n**Added**\n- New query sources: OpenAlex, PubMed, arXiv, dblp\n- Skill quality improvements: externalized prompt templates, added examples, fixed self-contained references\n\n### 0.1.0 (2026-05-27)\n\n- Initial release on PyPI\n\n---\n\n## 📄 License\n\nMIT License — see [LICENSE](LICENSE).\n","CiteCheck 是一个用于检查学术论文引用格式、可查询性、主题相关性和语义准确性的工具。其核心功能包括从 LaTeX 或 PDF 中提取参考文献，验证格式，并通过 Crossref、Semantic Scholar 等多个数据库验证引用的存在性，同时评估引用的文本相关性和语义匹配度，且作为技能使用时无需外部大语言模型 API 密钥。CiteCheck 既可作为跨代理技能支持 Claude Code、Codex 等多种 AI 编码助手，也提供独立的 Python 命令行接口供用户直接在终端运行。该工具特别适合需要确保学术写作中引用准确性与合规性的研究者和学生使用。",2,"2026-06-11 04:07:55","CREATED_QUERY"]