Citation Management
K-Dense-AI/claude-scientific-writer
Finds papers in OpenAlex, PubMed and Google Scholar, turns DOIs, PMIDs and arXiv IDs into clean BibTeX, and validates citations for a manuscript or thesis.
Exhaustively verify English and Chinese manuscript references before journal submission.
$ npx skills add Liuxiangjian-ai/reference-checker-skill --skill reference-checker -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install Liuxiangjian-ai/reference-checker-skill reference-checker --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
Claude Code skills documentation · loads skills from .claude/skills/
Install the "reference-checker" agent skill from https://github.com/Liuxiangjian-ai/reference-checker-skill/tree/main into .claude/skills/reference-checker/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "reference-checker", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add Liuxiangjian-ai/reference-checker-skill --skill reference-checker -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install Liuxiangjian-ai/reference-checker-skill reference-checker --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "reference-checker" agent skill from https://github.com/Liuxiangjian-ai/reference-checker-skill/tree/main into .agents/skills/reference-checker/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "reference-checker", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add Liuxiangjian-ai/reference-checker-skill --skill reference-checker -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install Liuxiangjian-ai/reference-checker-skill reference-checker --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "reference-checker" agent skill from https://github.com/Liuxiangjian-ai/reference-checker-skill/tree/main into .cursor/skills/reference-checker/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "reference-checker", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add Liuxiangjian-ai/reference-checker-skill --skill reference-checker -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install Liuxiangjian-ai/reference-checker-skill reference-checker --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "reference-checker" agent skill from https://github.com/Liuxiangjian-ai/reference-checker-skill/tree/main into .gemini/skills/reference-checker/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "reference-checker", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install Liuxiangjian-ai/reference-checker-skill reference-checkerInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add Liuxiangjian-ai/reference-checker-skill --skill reference-checker -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "reference-checker" agent skill from https://github.com/Liuxiangjian-ai/reference-checker-skill/tree/main into .github/skills/reference-checker/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "reference-checker", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add Liuxiangjian-ai/reference-checker-skill --skill reference-checker -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install Liuxiangjian-ai/reference-checker-skill reference-checker --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "reference-checker" agent skill from https://github.com/Liuxiangjian-ai/reference-checker-skill/tree/main into .opencode/skills/reference-checker/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "reference-checker", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
reference-checkerExhaustively verify English and Chinese manuscript references before journal submission.
Reference Checker is an agent skill from Liuxiangjian-ai/reference-checker-skill. Exhaustively verify English and Chinese manuscript references before journal submission. Use when checking whether references are real, accurate, complete, traceable, and formatted consistently. Uses DOI/PubMed/Crossref/publisher checks for English/international references, and uses Chinese-title-first CNKI/Wanfang/VIP/official-source checks for Chinese references. DOI is optional for Chinese references and must not be required unless the target citation style explicitly requires it.
Its SKILL.md is about 5.4k tokens, which your agent loads only when the skill is triggered. The skill folder holds 9 other files (for example `README.en.md`, `README.md` and `examples/expected_output.md`).
It sits in Research & Science, covering Citation management, Academic paper search and Scientific writing. It works with PubMed. The repository describes itself as: 逐条检查参考文献是否真实 /An exhaustive pre-submission reference verification skill for checking scholarly references item by item. The licence is MIT.
9 steps, taken from the first numbered list in SKILL.md.
Read from SKILL.md and the folder at commit f30bd18. It shows what the files ask for, not the result of running them.
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
No scripts in the folder and no shell commands in SKILL.md.
From the folder's file list and the shell code blocks in SKILL.md.
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Reference Checker loads about 5.4k tokens when it runs. Until then it costs about 127 tokens; SKILL.md has 2,730 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
The full file from Liuxiangjian-ai/reference-checker-skill at commit f30bd18, republished under its MIT licence (© Liuxiangjian-ai). 2,730 words, ~5,372 tokens.
.claude/skills/reference-checker/SKILL.md (or your agent's skills folder). This skill also uses 7 other files; get the full folder from GitHub.You are a pre-submission reference verification assistant. Your task is to audit every reference in a manuscript reference list for authenticity, bibliographic accuracy, DOI/PMID traceability for English/international sources, Chinese-title-first CNKI/Wanfang/VIP/official-source traceability for Chinese sources, duplication, source-type risks, and formatting consistency.
This skill is designed for exhaustive reference checking, not sampling.
It supports both English-language and Chinese-language references, including journal articles, books, dissertations, conference papers, standards, policies, reports, preprints, datasets, webpages, and other citable sources.
Unless the user explicitly requests sampling, you must check every reference one by one.
You must not:
If the reference list is too long for one response, process it in sequential batches. Continue from the last checked reference number in the next round.
For every reference, extract and display the following fields whenever available:
The audit table must include enough bibliographic information for the user to identify the reference without going back to the original list.
For each reference:
Parse the submitted reference into structured fields:
Identify the source language and source type:
Apply the correct verification route according to language and source type.
For English-language or international journal references:
For Chinese-language references:
DOI handling rules.
If the primary route fails, broaden by title and source.
Compare submitted metadata against verified metadata:
Assign a match quality:
Assign a status and severity:
Check whole-list issues:
Chinese references are verifiable even when they do not have DOI, PMID, or English metadata. Do not automatically downgrade them because they lack international identifiers.
For Chinese references following GB/T 7714 or common Chinese journal reference styles, DOI is normally optional unless the target journal explicitly asks for it. The primary verification basis should be Chinese title/source metadata from CNKI, Wanfang, VIP, official journal pages, or other authoritative Chinese sources, not DOI presence.
Use the following route whenever possible:
If CNKI and Wanfang disagree on minor formatting, prefer the official journal page when available. If no official page is accessible, record the discrepancy and mark confidence accordingly.
When searching Chinese references:
When a Chinese reference is verified through Chinese databases, record the database route clearly:
If the database page cannot be fully opened but search-result metadata is sufficient to identify the source, mark as Partial or Near exact depending on field completeness, and set confidence to Medium unless the official journal page confirms it.
For Chinese journal articles, verify:
Common issues to flag:
For Chinese dissertations, verify:
If a dissertation appears only in a university repository or only in CNKI/Wanfang, it can still be verified. Mark as Verified if identity fields match reliable repository metadata.
For Chinese books, verify:
Preferred verification routes:
For book chapters, additionally verify chapter title, chapter author, editor, page range, and book title.
For Chinese conference papers, verify:
If only an abstract or meeting report is found, mark as Partial or Manual check depending on citation style requirements.
For standards, policies, laws, regulations, and official documents, verify using official sources whenever possible.
Verify:
Preferred routes:
Flag superseded or abolished standards prominently.
For Chinese patents, verify:
Mark as Major if the patent status, applicant, or title is wrong. Mark as Critical if the number points to a different patent.
For English references, use the following preferred routes:
For biomedical references, PubMed, DOI metadata, and publisher records are preferred over secondary databases.
For each scholarly reference, check for visible retraction or expression-of-concern signals when feasible.
Flag:
Do not overstate retraction status unless a reliable source confirms it.
Every response must include a per-reference audit table for the references checked in that round.
Use this table format:
| Ref | Submitted Title | Submitted Authors | Submitted Source / Journal | Year | Identifier / Chinese Trace | Verification Route | Match Quality | Status | Confidence | Main Issue / Suggested Fix |
|---|
Rules for the table:
At the end of every response, state exactly which reference numbers were checked in this round.
If references remain unchecked, end with a continuation prompt:
"This round checked references X–Y. References Z–N remain unchecked. Would you like me to continue with the next round, references Z–...?"
Do not imply that the full audit is complete if references remain unchecked.
When all references have been checked, state clearly:
"Reference audit complete."
Then provide a cumulative summary of all problematic references found across the entire audit, grouped by severity:
| Ref | Title | Authors | Source / Journal | Problem | Suggested Action |
| Ref | Title | Authors | Source / Journal | Problem | Suggested Action |
| Ref | Title | Authors | Source / Journal | Problem | Suggested Action |
| Ref | Title | Authors | Source / Journal | Reason | Suggested Manual Search |
If no problems are found in a category, write "None identified."
Be concise but complete. Prioritize traceable verification over polished prose. Never fabricate metadata. If uncertain, say so clearly and mark the reference as Manual check.
For Chinese references, prefer clear Chinese wording in the issue/fix column. Preserve original Chinese names and titles unless the user asks for translated output.
© Liuxiangjian-ai, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
SKILL.md and 7 other files in the repository root of Liuxiangjian-ai/reference-checker-skill.
Open the folder on GitHubat commit f30bd18
Reference Checker next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Reference Checker this skillLiuxiangjian-ai/reference-checker-skill | 140 | — | ~5.4k | Automated safety check: Pass | MIT | |
| Citation ManagementK-Dense-AI/claude-scientific-writer | 2.4k | 2 repos | ~3.9k | Automated safety check: Notes | MIT | |
| Citation Managementforyourhealth111-pixel/Vibe-Skills | 3.6k | — | ~7.6k | Automated safety check: Notes | MIT | |
| Literature Reviewneflibata-feng/MyArxiv-Agent | 126 | 20 repos | ~5.9k | Automated safety check: Notes | MIT | |
| Citation Managementneflibata-feng/MyArxiv-Agent | 126 | 19 repos | ~8.1k | Automated safety check: Notes | MIT | |
| Autonomous Researchfedericodeponte/opendraft | 507 | — | ~8.2k | Automated safety check: Pass | Apache-2.0 |
K-Dense-AI/claude-scientific-writer
Finds papers in OpenAlex, PubMed and Google Scholar, turns DOIs, PMIDs and arXiv IDs into clean BibTeX, and validates citations for a manuscript or thesis.
foryourhealth111-pixel/Vibe-Skills
Turns DOIs, PMIDs and arXiv IDs into clean BibTeX, searches Google Scholar and PubMed, and checks and deduplicates a reference list.
neflibata-feng/MyArxiv-Agent
Conduct comprehensive, systematic literature reviews using multiple academic databases (PubMed, arXiv, bioRxiv, Semantic Scholar, etc.).
neflibata-feng/MyArxiv-Agent
Comprehensive citation management for academic research. An agent skill from neflibata-feng/MyArxiv-Agent.
federicodeponte/opendraft
An 18-agent pipeline that turns one topic line into a drafted research paper, literature review, or thesis chapter.
huggingface/skills
Indexes research papers on the Hugging Face Hub from arXiv, links them to models and datasets, claims authorship and generates markdown research articles from templates.
Works with
Categories
Exhaustively verify English and Chinese manuscript references before journal submission. Reference Checker is an agent skill from Liuxiangjian-ai/reference-checker-skill. Exhaustively verify English and Chinese manuscript references before journal submission.
Reference Checker fits situations like: checking whether references are real; formatted consistently.
Run `npx skills add Liuxiangjian-ai/reference-checker-skill --skill reference-checker -a claude-code`. Or copy the skill folder (the Liuxiangjian-ai/reference-checker-skill repository) into .claude/skills/reference-checker in your project. Claude Code loads it when a task matches its description.
Run `npx skills add Liuxiangjian-ai/reference-checker-skill --skill reference-checker -a codex`. Or copy the skill folder (the Liuxiangjian-ai/reference-checker-skill repository) into .agents/skills/reference-checker in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add Liuxiangjian-ai/reference-checker-skill --skill reference-checker -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/reference-checker, .gemini/skills/reference-checker, .github/skills/reference-checker and .opencode/skills/reference-checker in your project.
SKILL.md names no scripts, command-line tools or credentials: Reference Checker is instructions for the agent only.
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
Reference Checker is published under the MIT licence (from the LICENSE file in the skill folder). It allows redistribution, so the full SKILL.md is shown on this page.
About 5.4k tokens (SKILL.md is roughly 21k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
Skills that share tags, products or a category with Reference Checker: Citation Management (K-Dense-AI/claude-scientific-writer, 2.4k stars), Citation Management (foryourhealth111-pixel/Vibe-Skills, 3.6k stars), Literature Review (neflibata-feng/MyArxiv-Agent, 126 stars) and Citation Management (neflibata-feng/MyArxiv-Agent, 126 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Liuxiangjian-ai (a GitHub user) maintains it in Liuxiangjian-ai/reference-checker-skill, which has 140 GitHub stars. The repository was last updated on June 3, 2026.
Source: Liuxiangjian-ai/reference-checker-skill on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.