{
  "markdown": "# Scientific Skills for Claude\n\n![Demo](demo.gif)\n\nNatural language scientific literature search skills with semantic search over PubMed, arXiv, ChEMBL, DrugBank, bioRxiv, medRxiv, clinical trials, and more with zero API complexity.\n\n## Why These Skills Are Powerful\n\n- **No API Parameter Parsing**: Just pass natural language queries - no complex search syntax or parameters needed\n- **Semantic Search**: Understands meaning and context, not just keyword matching\n- **Full-Text Access**: Returns complete article content, not just abstracts\n- **Image Links**: Includes figures and images from papers\n- **Comprehensive Coverage**: Access to ALL scientific literature across multiple databases\n\n## Example\n\n![Example 1](example1.png)\n\n![Example 2](example2.png)\n\n## Available Skills\n\n### Individual Data Sources (9 skills)\n\n| Skill | Data Source | Coverage |\n|-------|-------------|----------|\n| `pubmed-search` | PubMed | Biomedical and life sciences literature |\n| `arxiv-search` | arXiv | Physics, mathematics, computer science, quantitative biology preprints |\n| `biorxiv-search` | bioRxiv | Biology preprints |\n| `medrxiv-search` | medRxiv | Medical and health sciences preprints |\n| `chembl-search` | ChEMBL | Bioactive molecules with drug-like properties |\n| `drugbank-search` | DrugBank | Comprehensive drug and drug target database |\n| `clinical-trials-search` | ClinicalTrials.gov | Clinical trials registry |\n| `drug-labels-search` | FDA Drug Labels | Official FDA drug information |\n| `open-targets-search` | Open Targets | Drug targets and disease associations |\n| `patents-search` | Patent Databases | Global patent filings |\n\n### Aggregated Skills (3 skills)\n\n| Skill | Combined Sources | Best For |\n|-------|------------------|----------|\n| `literature-search` | PubMed + arXiv + bioRxiv + medRxiv | General scientific literature review |\n| `biomedical-search` | PubMed + bioRxiv + medRxiv + ClinicalTrials + Drug Labels | Medical and clinical research |\n| `drug-discovery-search` | ChEMBL + DrugBank + Drug Labels + Open Targets | Drug discovery and development |\n\n## Installation\n\n### Via skills.sh (Recommended)\n\n```bash\nnpx skills i yorkeccak/scientific-skills\n```\n\nOr browse individual skills at [skills.sh](https://skills.sh) - search for \"yorkeccak\"\n\nExample: [skills.sh/yorkeccak/scientific-skills/biomedical-search](https://skills.sh/yorkeccak/scientific-skills/biomedical-search)\n\n**Supports installation to multiple agents:**\n- Antigravity\n- Claude Code\n- Clawdbot\n- Codex\n- Cursor\n- Droid\n- Gemini CLI\n- GitHub Copilot\n- Goose\n- Kilo Code\n- Kiro CLI\n- OpenCode\n- Roo Code\n- Trae\n- Windsurf\n\n### Via Claude Plugin\n\n```bash\n/plugin install yorkeccak/scientific-skills\n```\n\n### Manual Installation\n\n```bash\n# Clone the repository\ncd ~/dev\ngit clone https://github.com/yorkeccak/scientific-skills.git\n\n# Add as a local plugin\n/plugin add ~/dev/scientific-skills\n```\n\n## Quick Start\n\n### 1. Get an API Key for the Valyu search\n\nThere is $10 free credits: [platform.valyu.ai](https://platform.valyu.ai)\n\n### 2. First Use (Automatic Setup)\n\nJust start using any skill! Claude will automatically:\n1. Detect that no API key is configured\n2. Ask you to paste your API key\n3. Save it to `~/.valyu/config.json`\n4. Run your search\n\n```\nYou: Search PubMed for CRISPR advances in 2024\nClaude: \"To search PubMed, I need your Valyu API key from https://platform.valyu.ai\"\nYou: val_abc123...\nClaude: [saves key and runs search automatically]\n```\n\n### 3. Manual Setup (Optional)\n\nIf you prefer to set up manually:\n\n**Environment Variable** (recommended):\n```bash\n# For Zsh (macOS default)\necho 'export VALYU_API_KEY=\"your-api-key-here\"' >> ~/.zshrc\nsource ~/.zshrc\n\n# For Bash\necho 'export VALYU_API_KEY=\"your-api-key-here\"' >> ~/.bashrc\nsource ~/.bashrc\n```\n\n**Config File**:\n```bash\nmkdir -p ~/.valyu\necho '{\"apiKey\": \"your-api-key-here\"}' > ~/.valyu/config.json\n```\n\n\n## Output Format\n\nAll skills return consistent JSON output:\n\n```json\n{\n  \"success\": true,\n  \"type\": \"pubmed_search\",\n  \"query\": \"your query\",\n  \"result_count\": 10,\n  \"results\": [\n    {\n      \"title\": \"Article Title\",\n      \"url\": \"https://...\",\n      \"content\": \"Full article text with figures...\",\n      \"source\": \"pubmed\",\n      \"relevance_score\": 0.95,\n      \"images\": [\"https://example.com/figure1.jpg\"]\n    }\n  ],\n  \"cost\": 0.025\n}\n```\n\n## Processing Results with jq\n\n```bash\n# Get article titles\nscripts/search \"query\" 10 | jq -r '.results[].title'\n\n# Get URLs\nscripts/search \"query\" 10 | jq -r '.results[].url'\n\n# Extract content\nscripts/search \"query\" 10 | jq -r '.results[].content'\n\n# Filter by relevance score\nscripts/search \"query\" 20 | jq '.results[] | select(.relevance_score > 0.9)'\n```\n\n## Common Use Cases\n\n### Literature Review\n\nSearch across all scientific literature:\n\n```bash\ncd ~/dev/scientific-skills/skills/literature-search\nscripts/search \"epigenetic modifications in cancer\" 50\n```\n\n### Drug Discovery\n\nFind compounds and targets:\n\n```bash\ncd ~/dev/scientific-skills/skills/drug-discovery-search\nscripts/search \"kinase inhibitors for melanoma\" 30\n```\n\n### Clinical Research\n\nFind trials and patient outcomes:\n\n```bash\ncd ~/dev/scientific-skills/skills/biomedical-search\nscripts/search \"checkpoint inhibitor combinations trials\" 40\n```\n\n### Patent Research\n\nSearch for prior art:\n\n```bash\ncd ~/dev/scientific-skills/skills/patents-search\nscripts/search \"CRISPR delivery methods\" 25\n```\n\n### Gene Function Research\n\nUnderstand gene roles:\n\n```bash\ncd ~/dev/scientific-skills/skills/pubmed-search\nscripts/search \"TP53 tumor suppression mechanisms\" 20\n```\n\n## Requirements\n\n- Node.js 18+ (uses built-in fetch API)\n- Valyu API key from [platform.valyu.ai](https://platform.valyu.ai)\n- No npm packages required - zero external dependencies\n\n## Technical Details\n\n### Architecture\n\nEach skill consists of:\n```\nskill-name-search/\n├── SKILL.md              # Skill documentation\n└── scripts/\n    ├── search            # Bash wrapper\n    └── search.mjs        # Node.js implementation\n```\n\n## FAQ\n\n### What's the difference between individual and aggregated skills?\n\nIndividual skills search a single data source (e.g., only PubMed). Aggregated skills search multiple related sources simultaneously for more comprehensive results.\n\n### How is this different from regular PubMed search?\n\n- **Semantic understanding**: Understands query meaning, not just keywords\n- **Full-text access**: Returns complete articles, not just abstracts\n- **Natural language**: No need to learn search syntax\n- **Unified interface**: Same approach across all scientific databases\n\n### Can I customize the search parameters?\n\nThe skills use natural language queries by design to eliminate API complexity. For advanced filtering, you can:\n1. Be more specific in your natural language query\n2. Adjust the `maxResults` parameter\n3. Filter results using jq after receiving them\n\n## Comparison to Traditional Search\n\n| Feature | Traditional Search | These Skills |\n|---------|-------------------|--------------|\n| Query Format | Complex syntax, Boolean operators | Natural language |\n| Search Type | Keyword matching | Semantic understanding |\n| Content Returned | Abstracts, metadata | Full-text with images |\n| API Complexity | High - learn each API | Zero - same interface |\n| Setup Time | Hours per database | Minutes for all |\n| Multiple Sources | Write separate integrations | Single unified approach |\n\n## Contributing\n\nWelcome all PRs!\n\n## License\n\nMIT License - See individual skill files for details.\n",
  "bytes": 7478,
  "sha": "1a83ef0da34616547c99477de9c143729342078edef14510c1373d38846fffc6",
  "repo_slug": "yorkeccak/scientific-skills",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/skl_yorkeccak_scientific_skills_arxiv_search_9ffb6bef/readme"
}