メニュー
Produce a BibTeX/biblatex entry from an arXiv ID or DOI. Optionally appends to Paper/references.bib if it exists. Use when the user says "cite this", "bibtex for arXiv:...", "DOI 10.xxx/...", "add citation", or when adding a reference during paper writing.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Build an LLM-generated "dev notebook" for the paclet under development: a self-contained, evaluated, cloud-deployed Wolfram notebook that presents every exported function (a reference card) plus a sequence of worked examples, each mostly code and each producing a picture. Outputs are embedded (graphics rasterized, small symbolic results kept as live boxes), the build is smoke- tested headless to zero messages, deployed public to the Wolfram Cloud, and the stable URL is linked from the README. Use when the user says "dev notebook", "demo notebook", "build the paclet demo", "show what the paclet does", "reference notebook for the paclet", or the /dev-notebook command.
Create or modify Wolfram Notebooks (.nb) from structured Markdown content using the Wolfram MCP. This is the unified notebook skill — use it for creating new notebooks, editing existing ones, or converting NotebooksLLM/ markdown sources into .nb files. Triggers on: "create notebook", "make a notebook", "notebook about X", "edit notebook", "update notebook", "put this in a notebook", "generate .nb". Also used by other skills (new-project, start-tour) when they produce notebooks.
Add a resource (paper, repo, notebook, dataset, tool, web page) to the wiki. Downloads the file locally (gitignored), creates a wiki article with a Recover section for reproducibility, and updates the index. Use when the user says "add this paper", "save this resource", "note this repo", or when encountering a new reference during work. Also use during new-project to register discovered papers and Wolfram Community resources.
Build a Wolfram paclet archive and install it locally. Use when the user says "build paclet", "build the paclet", "install paclet locally", "make paclet", "create paclet archive", or "test the paclet install".
Run a health check on the wiki. Finds stale articles, missing coverage, broken backlinks, and opportunities for cross-linking. Use at the start of a new session, or when the user says "check wiki", "wiki health", "audit wiki", "wiki status".
Initialize a plain-markdown knowledge base (Wiki/) in the current repo. Use when the user says "init wiki", "set up knowledge base", "add wiki to this repo", or when starting a new project that needs a wiki. Also triggers when update-wiki or check-wiki is invoked but Wiki/ does not exist yet.
| name | cite |
| description | Produce a BibTeX/biblatex entry from an arXiv ID or DOI. Optionally appends to Paper/references.bib if it exists. Use when the user says "cite this", "bibtex for arXiv:...", "DOI 10.xxx/...", "add citation", or when adding a reference during paper writing. |
Thin wrapper around scripts/cite_from_id.wls that takes an identifier (arXiv ID, DOI, or URL containing either) and emits a BibTeX entry.
Optionally appends to Paper/references.bib and creates a wiki article via add-resource.
| Input | Example |
|---|---|
| New-style arXiv ID | 2301.00001 |
| Old-style arXiv ID | math-ph/0501001 |
| arXiv URL | https://arxiv.org/abs/2301.00001 |
arXiv: prefixed | arXiv:2301.00001 |
| Bare DOI | 10.1088/0951-7715/4/2/002 |
| DOI URL | https://doi.org/10.1088/... |
doi: prefixed | doi:10.1088/... |
cite_from_id.wls spawns a fresh wolframscript kernel (a single Import of the arXiv/Crossref metadata), which consumes a license seat.
Before running it, check headroom on the AgentTools MCP (no extra seat):
With[{free = $MaxLicenseProcesses - $LicenseProcesses}, free]
If free <= 0, do not spawn wolframscript — run the same Import[...] through mcp__Wolfram__WolframLanguageEvaluator instead.
The script remains the path when the MCP is unavailable.
wolframscript -f "${CLAUDE_PLUGIN_ROOT}/scripts/cite_from_id.wls" <identifier>
Output is a BibTeX entry on stdout.
For arXiv inputs the entry is built from the arXiv Atom API and includes title, author, year, eprint, archivePrefix, url, and abstract.
For DOI inputs the entry is fetched from Crossref's application/x-bibtex content negotiation.
Always show the raw BibTeX before writing anywhere. The user may want to edit the key, trim the abstract, or fix author casing.
Ask the user (or infer from context):
Paper/references.bib — if a Paper/ directory exists and the citation is for the paper being written.
Append at end, after checking the key is not already present (grep first).Journal/references.bib — if a Journal/ directory exists and the citation supports a journal entry (see the journal skill).
Append at end, after grep-checking the key is not already present, as for Paper.Wiki/Resources/<Key>.md — if the citation deserves a full wiki article (will be summarised).
Invoke add-resource with the BibTeX and the URL.If the user supplies multiple identifiers (newline- or comma-separated), run the script once per identifier, collect all BibTeX entries, deduplicate keys (suffix a, b, ... if collisions occur), then write the batch.
% arXiv fetch failed for <id>.
Do not write a broken entry.
Tell the user the ID couldn't be resolved.% Crossref fetch failed for <doi>.
Same: do not write.% Could not classify identifier.
Ask the user for a corrected ID.Paper/references.bib this skill appends to.Journal/references.bib; cite into it the same way as Paper.