Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →31 database search skills. Trigger: finding papers, search strategies, querying academic databases. Design: one skill per database/tool with API details, query syntax, and rate limits.
.claude/skills/brycewang-stanford-search-skills/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -19% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 252% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -16% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 13% | 0% |
Select the skill matching the user's need, then read its SKILL.md.
| Skill | Description | |-------|-------------| | arxiv-api | Search and retrieve preprints from the arXiv open-access repository | | arxiv-batch-reporting | Batch search and report generation from arXiv preprint repository | | arxiv-cli-tools | Command-line tools for searching and batch-downloading arXiv papers | | arxiv-paper-processor | Process and analyze arXiv papers systematically for research workflows | | baidu-scholar-guide | Using Baidu Scholar for Chinese and English academic literature search | | base-academic-search | Search 400M+ open access documents via the BASE search engine API | | biorxiv-api | Preprint server API for biology and medicine papers | | boolean-search-guide | Master Boolean operators and advanced search syntax for academic databases | | chatpaper-guide | Use ChatPaper to summarize and search arXiv papers with LLM assistance | | citation-chaining-guide | Forward and backward citation chaining techniques for literature search | | citeseerx-api | Search computer science literature via the CiteSeerX digital library | | database-comparison-guide | Compare major academic databases and when to use each for research | | deep-literature-search | Multi-source exhaustive literature search across academic databases | | eric-education-api | Search 2M+ education research records via the ERIC database API | | europe-pmc-api | Search biomedical and life sciences literature via Europe PMC | | findpapers-guide | Search multiple academic databases simultaneously with Findpapers | | google-scholar-guide | Advanced Google Scholar search techniques for comprehensive literature discovery | | ieee-xplore-api | Search IEEE's 6M+ engineering and CS publications via the Xplore API | | lens-scholarly-api | Search 300M+ scholarly and patent records via the Lens.org API | | mesh-terms-guide | Navigate MeSH vocabulary for precise PubMed and MEDLINE searches | | open-library-api | Search and access book metadata via the Open Library API | | open-semantic-search-guide | Self-hosted semantic search and text mining platform | | openaire-api | Search EU-funded research outputs via the OpenAIRE Graph API | | openalex-api | Query the OpenAlex catalog of scholarly works, authors, and institutions | | plos-open-access-api | Search PLOS open access journals with full-text Solr-powered API | | pubmed-api | Search biomedical literature and retrieve records via PubMed E-utilities | | scielo-api | Access Latin American and developing world research via SciELO API | | semantic-scholar-api | Search papers and analyze citation graphs via OpenAlex and CrossRef APIs | | share-research-api | Discover open access research outputs via the SHARE notification API | | systematic-search-strategy | Construct rigorous systematic search strategies for literature reviews | | worldcat-search-api | Search the world's largest library catalog via OCLC WorldCat API |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-05 | pass→fail | 14,092 | 3,664 | -74% | 1 | 1 | 0% | 2,070 | 1,258 | -39% | 0 | 0 | — |
case-19 | fail→fail | 9,377 | 4,748 | -49% | 1 | 1 | 0% | 1,415 | 1,309 | -7% | 0 | 0 | — |
case-01 | fail→pass | 19,237 | 16,033 | -17% | 1 | 1 | 0% | 2,878 | 2,319 | -19% | 0 | 0 | — |
case-02 | fail→pass | 3,118 | 3,529 | +13% | 1 | 1 | 0% | 445 | 1,565 | +252% | 0 | 0 | — |
case-03 | fail→pass | 12,132 | 4,295 | -65% | 1 | 1 | 0% | 1,953 | 1,737 | -11% | 0 | 0 | — |
case-04 | fail→pass | 11,464 | 17,395 | +52% | 1 | 1 | 0% | 1,995 | 1,675 | -16% | 0 | 0 | — |
case-06 | pass→pass | 9,620 | 2,760 | -71% | 1 | 1 | 0% | 1,383 | 1,392 | +1% | 0 | 0 | — |
case-07 | pass→pass | 11,599 | 2,990 | -74% | 1 | 1 | 0% | 1,923 | 1,316 | -32% | 0 | 0 | — |
case-08 | fail→pass | 12,518 | 6,448 | -48% | 1 | 1 | 0% | 1,919 | 2,159 | +13% | 0 | 0 | — |
case-09 | fail→pass | 16,382 | 3,171 | -81% | 1 | 1 | 0% | 2,636 | 1,454 | -45% | 0 | 0 | — |
case-10 | pass→pass | 12,201 | 4,170 | -66% | 1 | 1 | 0% | 1,832 | 1,409 | -23% | 0 | 0 | — |
case-11 | pass→fail | 13,329 | 4,448 | -67% | 1 | 1 | 0% | 2,211 | 1,308 | -41% | 0 | 0 | — |
case-12 | pass→fail | 10,046 | 5,602 | -44% | 1 | 1 | 0% | 1,542 | 1,241 | -20% | 0 | 0 | — |
case-13 | fail→pass | 12,447 | 2,143 | -83% | 1 | 1 | 0% | 1,766 | 1,349 | -24% | 0 | 0 | — |
case-14 | fail→pass | 9,318 | 5,289 | -43% | 1 | 1 | 0% | 1,417 | 1,441 | +2% | 0 | 0 | — |
case-15 | pass→pass | 11,385 | 3,150 | -72% | 1 | 1 | 0% | 1,785 | 1,431 | -20% | 0 | 0 | — |
case-16 | pass→fail | 22,035 | 2,839 | -87% | 1 | 1 | 0% | 3,093 | 1,254 | -59% | 0 | 0 | — |
case-17 | fail→pass | 12,191 | 2,083 | -83% | 1 | 1 | 0% | 1,955 | 1,350 | -31% | 0 | 0 | — |
case-18 | fail→fail | 9,391 | 4,658 | -50% | 1 | 1 | 0% | 1,439 | 1,272 | -12% | 0 | 0 | — |
case-20 | pass→fail | 9,456 | 11,138 | +18% | 1 | 1 | 0% | 1,610 | 1,256 | -22% | 0 | 0 | — |
case-21 | pass→fail | 8,626 | 2,384 | -72% | 1 | 1 | 0% | 1,382 | 1,248 | -10% | 0 | 0 | — |
case-22 | pass→fail | 9,315 | 15,005 | +61% | 1 | 1 | 0% | 1,485 | 1,227 | -17% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 13 counted toward the lift figure. The other 9 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +9 percentage points is the difference between those two pass rates over the 13 comparable cases. 7 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.