菜单
Find academic papers, explore citations, and export BibTeX. Use when asked about research papers, references, citations, bibliography, or scholarly information.
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。
Multi-agent review-and-improve loop for a GitHub PR you have checked out — posts a "starting" PR comment cc'ing the original author, runs requested rounds plus any adaptive continuation, applies fix commits to the local branch after each round, pushes everything back to the PR, then edits the starting comment in-place with the synthesized report (or a failure summary). Auto-detects the PR from the currently checked-out branch when no locator is supplied. Use when the user wants to "improve a PR", "review and commit fixes", "iterate on my PR", or "review and push back" against a checked-out PR branch. Requires `gh`, `uuidgen`, `jq`, and `uv` or `python3` on PATH. Activates the `review-anvil` engine in per_fix mode.
Read-only multi-agent review of a GitHub Pull Request, with the synthesized report posted back as a PR comment so the author is notified. Use when the user wants to review a GitHub PR (github.com or GitHub Enterprise) and post a structured review back to the PR conversation. Auto-detects the PR from the currently checked-out branch when no locator is supplied. Requires `gh`, `uuidgen`, `jq`, and `uv` or `python3` on PATH. Activates the `review-anvil` engine in read-only mode and orchestrates the shell helper for posting.
Multi-agent code review without any edits or commits — requested rounds of parallel reviewers, synthesis, default-on reproduction of uncertain material findings, and a final report. Use when the user wants a code review of a diff, branch, path, or PR but does NOT want any fixes applied automatically (no edits, no commits, no staging). Activates the `review-anvil` engine in read-only mode.
Iteratively refine code via requested rounds of parallel subagent review and orchestrator-applied fixes, with bounded adaptive continuation enabled by default for productive runs. Use when the user says "let's do three rounds", "fix/review loop", "back-and-forth review", "iterative review", or asks to harden a change with multiple rounds of codex/claude review.
Convert academic PDFs to clean Markdown. Use whenever you encounter a PDF paper (local file or URL) that needs to be read, analyzed, or referenced as text.
Delegate code review, plan review, and exploration to Codex CLI. Use when you need an independent second opinion on code changes, plans, or architecture — or when asked to run codex review/exec.
| name | scholar-search |
| description | Find academic papers, explore citations, and export BibTeX. Use when asked about research papers, references, citations, bibliography, or scholarly information. |
Find papers with s2cli (Semantic Scholar), openalexcli (OpenAlex), and arxivy (arXiv), get high-quality BibTeX with dblpcli (DBLP). All tools run via uvx — no installation needed.
Three tools for paper discovery, each with different strengths.
Best for: paper recommendations, citation/reference networks, arXiv/DOI lookups, citation count filtering.
# Search by keyword
uvx s2cli search "attention mechanism" --limit 10 --json
uvx s2cli search "transformers" --year 2020-2024 --min-citations 100 --json
# Filter options: --year, --venue, --field, --min-citations, --open-access
# Get paper details by ID (supports S2 ID, DOI, arXiv, CorpusId)
uvx s2cli paper ARXIV:1706.03762 --json
uvx s2cli paper DOI:10.18653/v1/N18-3011 --json
# Citation network
uvx s2cli citations ARXIV:1706.03762 --limit 20 --json # papers citing this
uvx s2cli references ARXIV:1706.03762 --limit 20 --json # papers cited by this
# Recommendations based on a seed paper
uvx s2cli recommend ARXIV:1706.03762 --limit 10 --json
uvx s2cli recommend ARXIV:1706.03762 --pool all-cs --json # pool: 'recent' or 'all-cs'
# Author search and papers
uvx s2cli author search "Yoshua Bengio" --json
uvx s2cli author papers <author_id> --json
Best for: broadest coverage (260M+ works), institution/journal queries, group-by aggregation, ORCID lookups, date range filtering.
# Search by keyword
uvx openalexcli search "attention mechanism" --limit 10 --json
uvx openalexcli search "transformers" --from-date 2020-01-01 --min-citations 100 --json
# Filter options: --filter, --from-date, --to-date, --min-citations, --open-access, --type, --sort, --group-by
# Get work by OpenAlex ID
uvx openalexcli work W2741809807 --json
# Citation network
uvx openalexcli citations W2741809807 --limit 20 --json
uvx openalexcli references W2741809807 --limit 20 --json
# Author search and works
uvx openalexcli author search "Yoshua Bengio" --json
uvx openalexcli author works <author_id> --json
# Institution queries
uvx openalexcli institution search "MIT" --json
uvx openalexcli institution works <institution_id> --json
# Journal/source queries
uvx openalexcli source search "Nature" --json
uvx openalexcli source works <source_id> --json
# Aggregation (unique to OpenAlex)
uvx openalexcli search "deep learning" --group-by publication_year --json
Best for: searching arXiv directly, getting paper details by arXiv ID, browsing recent submissions in a category. No API key needed.
# Search by keyword
uvx arxivy search "attention mechanism" --limit 10
uvx arxivy search "transformers" -c cs.CL -n 20
uvx arxivy search "diffusion models" --sort submittedDate
# Include abstracts in results
uvx arxivy search "RLHF" -c cs.AI --abstract
# Get paper details by arXiv ID (accepts bare IDs, versioned IDs, or full URLs)
uvx arxivy paper 1706.03762
uvx arxivy paper 1706.03762 2010.11929 1810.04805 # compare multiple
uvx arxivy paper 1706.03762 --json
# Browse recent submissions in a category
uvx arxivy new cs.AI
uvx arxivy new cs.CL -n 20 --abstract
# BibTeX export (fallback — prefer dblpcli for curated entries)
uvx arxivy bibtex 1706.03762
uvx arxivy bibtex 1706.03762 2010.11929 > refs.bib
Note on output: arxivy auto-detects whether stdout is a terminal. In a terminal it shows Rich tables; when piped it outputs JSON. Use --json to force JSON in a terminal.
| Need | Use |
|---|---|
| Recommendations for a paper | s2cli recommend |
| Citation/reference graph traversal | s2cli citations / s2cli references |
| arXiv ID or DOI lookup | s2cli paper or arxivy paper |
| Search arXiv directly (by category, recency) | arxivy search |
| Browse new arXiv submissions | arxivy new |
| Broadest search coverage | openalexcli search |
| Filter by institution or journal | openalexcli institution / openalexcli source |
| Aggregate stats (group-by) | openalexcli search --group-by |
| ORCID-based author lookup | openalexcli author get |
Use dblpcli for high-quality BibTeX sourced directly from DBLP (curated, consistent formatting). If a paper isn't in DBLP (e.g. a very recent arXiv preprint), fall back to uvx arxivy bibtex <arxiv_id>.
Find the paper's DBLP key by searching:
uvx dblpcli search "attention is all you need" --format json
The key field in results (e.g., conf/nips/VaswaniSPUJGKP17) is what you need.
Export BibTeX:
uvx dblpcli bibtex conf/nips/VaswaniSPUJGKP17
Batch export (multiple keys):
uvx dblpcli bibtex conf/nips/VaswaniSPUJGKP17 journals/jmlr/KingmaB14
Custom citation key:
uvx dblpcli bibtex conf/nips/VaswaniSPUJGKP17 --key vaswani2017attention
Save to file:
uvx dblpcli bibtex conf/nips/VaswaniSPUJGKP17 --output refs.bib
# Search publications (supports AND/OR operators, author: prefix)
uvx dblpcli search "transformer attention" --format json
uvx dblpcli search "deep learning" --year 2020-2024 --format json
uvx dblpcli search "author:Vaswani" --limit 10 --format json
# Get publication details
uvx dblpcli pub conf/nips/VaswaniSPUJGKP17 --format json
# Author workflow
uvx dblpcli author search "Geoffrey Hinton" --format json # find PID
uvx dblpcli author get <pid> --format json # author details
uvx dblpcli author pubs <pid> --format json # list publications
uvx dblpcli author bibtex <pid> # BibTeX for ALL papers
# Venue queries
uvx dblpcli venue search "NeurIPS" --format json
uvx dblpcli venue pubs conf/nips --year 2023 --format json
uvx s2cli search "topic" --json or uvx openalexcli search "topic" --jsonuvx s2cli citations <id> --jsonuvx s2cli recommend <id> --jsonuvx dblpcli search "paper title" --format jsonuvx dblpcli bibtex <key1> <key2> <key3> --output refs.bibuvx dblpcli search "title" --format jsonuvx dblpcli bibtex <key1> <key2> ... --output refs.bibuvx s2cli paper ARXIV:1706.03762 --jsonuvx s2cli citations ARXIV:1706.03762 --limit 20 --jsonuvx s2cli references ARXIV:1706.03762 --limit 20 --jsonuvx s2cli recommend ARXIV:1706.03762 --jsonuvx dblpcli author search "Name" --format json → uvx dblpcli author bibtex <pid>uvx s2cli author search "Name" --json → uvx s2cli author papers <id> --jsonuvx openalexcli author search "Name" --json → uvx openalexcli author works <id> --json| Variable | Purpose | Required |
|---|---|---|
S2_API_KEY | Semantic Scholar API key (higher rate limits) | No |
OPENALEX_EMAIL | OpenAlex polite pool (faster responses) | No |
Both tools work without these but may hit rate limits on heavy use.
Use this skill when the user asks about: