$npx -y skills add coleam00/custom-agent-with-skills --skill research_assistantSearch academic papers, find citations, and explore research using Semantic Scholar. Use when user asks about scientific papers, research, citations, or academic topics.
| 1 | # Research Assistant Skill |
| 2 | |
| 3 | Search and explore academic research using the Semantic Scholar API. Access over 214 million papers, 2.49 billion citations, and 79 million authors across all fields of science. |
| 4 | |
| 5 | ## When to Use |
| 6 | |
| 7 | - User asks about scientific or academic papers |
| 8 | - User wants to find research on a topic |
| 9 | - User asks for citations or references |
| 10 | - User wants to understand the state of research in a field |
| 11 | - User asks "what does the research say about..." |
| 12 | - User wants to find papers by a specific author |
| 13 | - User needs to find papers related to another paper |
| 14 | |
| 15 | ## Available Operations |
| 16 | |
| 17 | 1. **Paper Search**: Find papers by keywords, title, or abstract |
| 18 | 2. **Author Lookup**: Find papers by a specific researcher |
| 19 | 3. **Citation Exploration**: Find papers that cite or are cited by a paper |
| 20 | 4. **Paper Recommendations**: Get papers similar to a given paper |
| 21 | 5. **Paper Details**: Get full metadata including abstract, authors, venue, year |
| 22 | |
| 23 | ## Instructions |
| 24 | |
| 25 | When a user asks about research or academic papers: |
| 26 | |
| 27 | ### Step 1: Understand the Query |
| 28 | |
| 29 | Identify what the user wants: |
| 30 | - **Topic search**: "papers about transformer architectures" |
| 31 | - **Author search**: "papers by Geoffrey Hinton" |
| 32 | - **Citation search**: "what papers cite this one" |
| 33 | - **Recommendations**: "papers similar to Attention Is All You Need" |
| 34 | |
| 35 | ### Step 2: Load API Reference |
| 36 | |
| 37 | ALWAYS read `references/api_reference.md` before making API calls to ensure you use the correct endpoints and parameters. |
| 38 | |
| 39 | ### Step 3: Construct the Search |
| 40 | |
| 41 | **IMPORTANT: Make ONE well-constructed search query.** Do not make multiple API calls with variations of the same query. The API has rate limits, and multiple calls waste quota. |
| 42 | |
| 43 | For topic searches, use the paper search endpoint: |
| 44 | ``` |
| 45 | https://api.semanticscholar.org/graph/v1/paper/search?query=YOUR_QUERY |
| 46 | ``` |
| 47 | |
| 48 | Tips for effective single queries: |
| 49 | - Use specific, focused keywords (e.g., "transformer attention mechanism" not just "transformer") |
| 50 | - Include relevant filters like `year`, `minCitationCount`, or `fieldsOfStudy` to narrow results |
| 51 | - Request enough fields in one call: `title,authors,year,citationCount,abstract,tldr` |
| 52 | - If the first query returns no results, ask the user to refine their search rather than trying variations |
| 53 | |
| 54 | Key parameters: |
| 55 | - `query`: Search terms (required) |
| 56 | - `fields`: Which data to return (see API reference) |
| 57 | - `limit`: Number of results (default 10, max 100) |
| 58 | - `offset`: For pagination |
| 59 | |
| 60 | ### Step 4: Present Results |
| 61 | |
| 62 | Format results clearly with: |
| 63 | - Paper title |
| 64 | - Authors (first author et al. if many) |
| 65 | - Year published |
| 66 | - Citation count |
| 67 | - Brief abstract or TLDR if available |
| 68 | - Link to paper (use paperId to construct URL) |
| 69 | |
| 70 | ### Common City Coordinates (for reference, similar to weather skill pattern) |
| 71 | |
| 72 | Not applicable - this skill uses text-based search queries. |
| 73 | |
| 74 | ### Paper ID Formats |
| 75 | |
| 76 | Semantic Scholar accepts multiple ID formats: |
| 77 | | Format | Example | |
| 78 | |--------|---------| |
| 79 | | Semantic Scholar ID | `649def34f8be52c8b66281af98ae884c09aef38b` | |
| 80 | | DOI | `DOI:10.18653/v1/N18-3011` | |
| 81 | | ArXiv | `ARXIV:2106.15928` | |
| 82 | | ACL | `ACL:W12-3903` | |
| 83 | | PubMed | `PMID:19872477` | |
| 84 | | Corpus ID | `CorpusId:215416146` | |
| 85 | |
| 86 | ## Resources |
| 87 | |
| 88 | ALWAYS read these before making API calls: |
| 89 | - `references/api_reference.md` - Complete Semantic Scholar API documentation |
| 90 | - `references/search_tips.md` - Advanced search techniques and query optimization |
| 91 | |
| 92 | ## Examples |
| 93 | |
| 94 | ### Example 1: Topic Search |
| 95 | User asks: "Find papers about large language models from 2023" |
| 96 | |
| 97 | 1. Load API reference |
| 98 | 2. Call: `https://api.semanticscholar.org/graph/v1/paper/search?query=large+language+models&year=2023&fields=title,authors,year,citationCount,abstract&limit=10` |
| 99 | 3. Format results with titles, authors, years, and citation counts |
| 100 | |
| 101 | ### Example 2: Author Search |
| 102 | User asks: "What has Yann LeCun published recently?" |
| 103 | |
| 104 | 1. First find author ID via author search |
| 105 | 2. Then get author's papers via author endpoint |
| 106 | 3. Filter/sort by year for recent work |
| 107 | |
| 108 | ### Example 3: Citation Analysis |
| 109 | User asks: "What papers cite 'Attention Is All You Need'?" |
| 110 | |
| 111 | 1. Find paper ID for "Attention Is All You Need" |
| 112 | 2. Use citations endpoint to get citing papers |
| 113 | 3. Present most influential citations (by citation count) |
| 114 | |
| 115 | ### Example 4: Research Overview |
| 116 | User asks: "What does the research say about sleep and memory?" |
| 117 | |
| 118 | 1. Search for papers on "sleep memory consolidation" |
| 119 | 2. Get top cited papers to find influential work |
| 120 | 3. Summarize key findings from abstracts |
| 121 | 4. Provide references for further reading |
| 122 | |
| 123 | ## Notes |
| 124 | |
| 125 | - The API is free and does not require authentication for basic use |
| 126 | - Rate limit: 1000 requests/second shared among unauthenticated users |
| 127 | - For heavy use, request an API key for dedicated rate limits |
| 128 | - Always include the `fields` parameter to get useful data back |
| 129 | - Papers include TLDRs (AI-generat |