ScienceClaw

Search Allen AI's Asta Scientific Corpus (225M+ papers, 12M+ full-text, 2.4B+ citations) via MCP endpoint. Provides paragraph-level semantic search across full-text publications, citation graph traversal, and author analysis. Use as a complement to PubMed/OpenAlex/Semantic Scholar for deeper literature discovery, especially when full-text search or citation network analysis is needed. Requires ASTA_API_KEY in .env (free registration at allenai.org/asta).

Browser automation for accessing scientific databases that lack REST APIs. Uses the browser-use Python framework (81k+ GitHub stars) to control a real browser via LLM vision. Enables data extraction from web-only databases like GEPIA2, GeneCards advanced features, COSMIC public data, and journal full-text access. Use as a fallback when curl-based API access fails or when the target database has no programmatic API. Requires pip install browser-use and a Chromium browser.

Systematic drug repurposing analysis inspired by NovusAI. Evaluates existing drugs for new therapeutic indications through multi-dimensional evidence gathering across target networks, clinical trials (including failures), patent landscape, safety profiles, and off-label literature. Produces ranked repurposing candidates with evidence scores. Use when users ask about finding new uses for existing drugs, off-label potential, "老药新用", or "drug repurposing for X". Complements target-validation (which starts from a target) by starting from a drug.

Evolving memory system inspired by EvoScientist. Extends ScienceClaw's research memory with four record types (finding, ideation, strategy, pitfall) to enable learning from past research sessions. Recall relevant strategies and pitfalls before recipe execution, extract and persist new lessons after completion. Use at the start and end of every research recipe, and when the user asks to recall past experience or improve workflows.

Five-step figure generation pipeline inspired by PaperVizAgent (Google Research, 2026). Orchestrates Retriever → Planner → Stylist → Visualizer → Critic stages for publication-quality scientific figures. Retrieves reference figures from literature, plans layout and composition, applies journal-specific styling, generates the figure, then critiques and refines. Use when the user needs high-quality figures for papers/presentations and wants a more deliberate, reference-driven approach than direct code generation. Especially useful for multi-panel figures and complex data visualizations.

AI-powered manuscript review and revision system inspired by APRES (ICLR 2026). Evaluates scientific manuscripts using ScholarEval 8-dimension rubric plus citation-predictive heuristics, then performs targeted revisions while preserving core scientific claims. Outputs before/after comparison with improvement metrics. Use when the user says "/review", "帮我审一下", "review my manuscript", "improve this paper", "polish this draft", or provides a manuscript for quality improvement. Also triggered by "审稿", "修改论文", "润色".

Autonomous molecular dynamics simulation pipeline inspired by DynaMate (2026). Designs, executes, and analyzes complete MD workflows for protein and protein-ligand systems. Covers structure retrieval, system preparation, minimization, equilibration, production, and trajectory analysis (RMSD, RMSF, hydrogen bonds, binding free energy). Uses OpenMM as the primary engine with AmberTools for preparation. Self-correcting — detects and fixes common simulation failures. Use when users ask for MD simulations, protein stability analysis, binding free energy calculations, or "跑个分子动力学模拟". Requires OpenMM and optionally AmberTools.

Pre-built research workflow templates that execute complete multi-step analyses from a single user prompt. Triggers on gene analysis, target validation, literature review, differential expression, clinical queries, researcher profiling, drug repurposing, or molecular dynamics simulation. Use when the user's query matches a Recipe pattern defined in SCIENCE.md.

Generate editable SVG scientific illustrations (mechanism diagrams, signaling pathways, workflow figures) directly from LLM text output. Uses a Review-Refine loop inspired by AutoFigure (ICLR 2026). Outputs editable SVG files compatible with draw.io, Illustrator, and PowerPoint, plus PNG renders. Use when the user needs mechanism diagrams, pathway illustrations, experimental workflow figures, or any schematic that should be editable. Complements the existing Gemini-based scientific-diagram-generation skill (which produces non-editable PNG).

Multi-step therapeutic reasoning inspired by TxAgent (Harvard MIMS, 2026). Chains ToolUniverse tool calls with clinical reasoning to analyze drug interactions, contraindications, and personalized treatment strategies. Goes beyond single-step database lookups to perform sequential reasoning across patient profile, pharmacology, guidelines, and safety data. Use when users ask about treatment recommendations, drug interactions for specific patients, personalized therapy, or "X 怎么治" with patient-specific context (age, comorbidities, concurrent medications). Enhances the existing clinical-query and target-validation recipes.

Export research project reports to Word (.docx) format with embedded figures and formatted references. Use when user says "导出 Word", "/export word", "转 docx", "生成 Word 报告", "export to Word", or wants a Word document from project results.

Export research project findings to a LaTeX manuscript draft with figures, references, and methods. Supports Nature, Cell, Lancet, and generic article formats. Use when user says "导出 LaTeX", "/export latex", "写论文初稿", "export to LaTeX", "generate manuscript", or wants a paper draft from project results. Builds on venue-templates skill.

Export research project findings to a presentation (.pptx) with key findings, figures, and conclusions. Use when user says "导出 PPT", "/export pptx", "做个汇报", "生成 PPT", "export to PowerPoint", "make a presentation from results", or wants slides from project results. Builds on pptx-generation skill.

Monitor research topics and alert the user when new papers are published. Use when user says "/watch", "监控", "关注这个课题", "有新文献告诉我", "monitor this topic", "alert me on new papers", "track new publications". Stores watch configurations and checks for new results at session start.

Write comprehensive clinical reports including case reports (CARE guidelines), diagnostic reports (radiology/pathology/lab), clinical trial reports (ICH-E3, SAE, CSR), and patient documentation (SOAP, H&P, discharge summaries). Full support with templates, regulatory compliance (HIPAA, FDA, ICH-GCP), and validation tools.

Create high-quality ToolUniverse skills following test-driven, implementation-agnostic methodology. Integrates tools from ToolUniverse's 1,264+ tool library, creates missing tools when needed using devtu-create-tool, tests thoroughly, and produces skills with Python SDK + MCP support. Use when asked to create new ToolUniverse skills, build research workflows, or develop domain-specific analysis capabilities for biology, chemistry, or medicine.

Automatically discover life science APIs online, create ToolUniverse tools, validate them, and prepare integration PRs. Performs gap analysis to identify missing tool categories, web searches for APIs, automated tool creation using devtu-create-tool patterns, validation with devtu-fix-tool, and git workflow management. Use when expanding ToolUniverse coverage, adding new API integrations, or systematically discovering scientific resources.

Optimize ToolUniverse skills for better report quality, evidence handling, and user experience. Apply patterns like tool verification, foundation data layers, disambiguation-first, evidence grading, quantified completeness, and report-only output. Use when reviewing skills, improving existing skills, or creating new ToolUniverse research skills.

Create HTML-based presentations that run in a browser. Use ONLY when the user explicitly asks for an HTML presentation or web-based slides. For PowerPoint (.pptx) generation, use the pptx-generation skill instead.

Create professional research posters in LaTeX using beamerposter, tikzposter, or baposter. Support for conference presentations, academic posters, and scientific communication. Includes layout design, color schemes, multi-column formats, figure integration, and poster-specific best practices for visual communication.

Generate comprehensive market research reports (50+ pages) in the style of top consulting firms (McKinsey, BCG, Gartner). Features professional LaTeX formatting, extensive visual generation with scientific-schematics and generate-image, deep integration with research-lookup for data gathering, and multi-framework strategic analysis including Porter Five Forces, PESTLE, SWOT, TAM/SAM/SOM, and BCG Matrix.

Optional integration with K-Dense Web for end-to-end multi-agent research workflows. Use when the user asks about K-Dense or needs complex research orchestration beyond single-agent capability.

Generate academic PowerPoint presentations (.pptx) using python-pptx. Use this skill for making PPT, slides, presentations, 生成PPT, 做PPT, 写PPT, 幻灯片. Provides complete helper functions and templates. Preferred over scientific-slides and frontend-slides for all PPTX generation.

Write competitive research proposals for NSF, NIH, DOE, DARPA, and Taiwan NSTC. Agency-specific formatting, review criteria, budget preparation, broader impacts, significance statements, innovation narratives, and compliance with submission requirements.

Slide design principles for scientific presentations — structure, pacing, visual hierarchy. For actual PPTX file generation, use the pptx-generation skill instead.

Core skill for the deep research and writing tool. Write scientific manuscripts in full paragraphs (never bullet points). Use two-stage process with (1) section outlines with key points using research-lookup then (2) convert to flowing prose. IMRAD structure, citations (APA/AMA/Vancouver), figures/tables, reporting guidelines (CONSORT/STROBE/PRISMA), for research papers and journal submissions.

Detect and analyze adverse drug event signals using FDA FAERS data, drug labels, disproportionality analysis (PRR, ROR, IC), and biomedical evidence. Generates quantitative safety signal scores (0-100) with evidence grading. Use for post-market surveillance, pharmacovigilance, drug safety assessment, adverse event investigation, and regulatory decision support.

Comprehensive antibody engineering and optimization for therapeutic development. Covers humanization, affinity maturation, developability assessment, and immunogenicity prediction. Use when asked to optimize antibodies, humanize sequences, or engineer therapeutic antibodies from lead to clinical candidate.

Discover novel small molecule binders for protein targets using structure-based and ligand-based approaches. Creates actionable reports with candidate compounds, ADMET profiles, and synthesis feasibility. Use when users ask to find small molecules for a target, identify novel binders, perform virtual screening, or need hit-to-lead compound identification.

Provide comprehensive clinical interpretation of somatic mutations in cancer. Given a gene symbol + variant (e.g., EGFR L858R, BRAF V600E) and optional cancer type, performs multi-database analysis covering clinical evidence (CIViC), mutation prevalence (cBioPortal), therapeutic associations (OpenTargets, ChEMBL, FDA), resistance mechanisms, clinical trials, prognostic impact, and pathway context. Generates an evidence-graded markdown report with actionable recommendations for precision oncology. Use when oncologists, molecular tumor boards, or researchers ask about treatment options for specific cancer mutations, resistance mechanisms, or clinical trial matching.

Comprehensive chemical safety and toxicology assessment integrating ADMET-AI predictions, CTD toxicogenomics, FDA label safety data, DrugBank safety profiles, and STITCH chemical-protein interactions. Performs predictive toxicology (AMES, DILI, LD50, carcinogenicity), organ/system toxicity profiling, chemical-gene-disease relationship mapping, regulatory safety extraction, and environmental hazard assessment. Use when asked about chemical toxicity, drug safety profiling, ADMET properties, environmental health risks, chemical hazard assessment, or toxicogenomic analysis.

Strategic clinical trial design feasibility assessment using ToolUniverse. Evaluates patient population sizing, biomarker prevalence, endpoint selection, comparator analysis, safety monitoring, and regulatory pathways. Creates comprehensive feasibility reports with evidence grading, enrollment projections, and trial design recommendations. Use when planning Phase 1/2 trials, assessing trial feasibility, or designing biomarker-driven studies.

AI-driven patient-to-trial matching for precision medicine and oncology. Given a patient profile (disease, molecular alterations, stage, prior treatments), discovers and ranks clinical trials from ClinicalTrials.gov using multi-dimensional matching across molecular eligibility, clinical criteria, drug-biomarker alignment, evidence strength, and geographic feasibility. Produces a quantitative Trial Match Score (0-100) per trial with tiered recommendations and a comprehensive markdown report. Use when oncologists, molecular tumor boards, or patients ask about clinical trial options for specific cancer types, biomarker profiles, or post-progression scenarios.

Comprehensive CRISPR screen analysis for functional genomics. Analyze pooled or arrayed CRISPR screens (knockout, activation, interference) to identify essential genes, synthetic lethal interactions, and drug targets. Perform sgRNA count processing, gene-level scoring (MAGeCK, BAGEL), quality control, pathway enrichment, and drug target prioritization. Use for CRISPR screen analysis, gene essentiality studies, synthetic lethality detection, functional genomics, drug target validation, or identifying genetic vulnerabilities.

Generates comprehensive drug research reports with compound disambiguation, evidence grading, and mandatory completeness sections. Covers identity, chemistry, pharmacology, targets, clinical trials, safety, pharmacogenomics, and ADMET properties. Use when users ask about drugs, medications, therapeutics, or need drug profiling, safety assessment, or clinical development research.

Comprehensive computational validation of drug targets for early-stage drug discovery. Evaluates targets across 10 dimensions (disambiguation, disease association, druggability, chemical matter, clinical precedent, safety, pathway context, validation evidence, structural insights, validation roadmap) using 60+ ToolUniverse tools. Produces a quantitative Target Validation Score (0-100) with GO/NO-GO recommendation. Use when users ask about target validation, druggability assessment, target prioritization, or "is X a good drug target for Y?"

Production-ready genomics and epigenomics data processing for BixBench questions. Handles methylation array analysis (CpG filtering, differential methylation, age-related CpG detection, chromosome-level density), ChIP-seq peak analysis (peak calling, motif enrichment, coverage stats), ATAC-seq chromatin accessibility, multi-omics integration (expression + methylation correlation), and genome-wide statistics. Pure Python computation (pandas, scipy, numpy, pysam, statsmodels) plus ToolUniverse annotation tools (Ensembl, ENCODE, SCREEN, JASPAR, ReMap, RegulomeDB, ChIPAtlas). Supports BED, BigWig, methylation beta-value matrices, Illumina manifest files, and multi-sample clinical data. Use when processing methylation data, ChIP-seq peaks, ATAC-seq signals, or answering questions about CpG sites, differential methylation, chromatin accessibility, histone marks, or epigenomic statistics.

Comprehensive immune repertoire analysis for T-cell and B-cell receptor sequencing data. Analyze TCR/BCR repertoires to assess clonality, diversity, V(D)J gene usage, CDR3 characteristics, convergence, and predict epitope specificity. Integrate with single-cell data for clonotype-phenotype associations. Use for adaptive immune response profiling, cancer immunotherapy research, vaccine response assessment, autoimmune disease studies, or repertoire diversity analysis in immunology research.

Predict patient response to immune checkpoint inhibitors (ICIs) using multi-biomarker integration. Given a cancer type, somatic mutations, and optional biomarkers (TMB, PD-L1, MSI status), performs systematic analysis across 11 phases covering TMB classification, neoantigen burden estimation, MSI/MMR assessment, PD-L1 evaluation, immune microenvironment profiling, mutation-based resistance/sensitivity prediction, clinical evidence retrieval, and multi-biomarker score integration. Generates a quantitative ICI Response Score (0-100), response likelihood tier, specific ICI drug recommendations with evidence, resistance risk factors, and a monitoring plan. Use when oncologists ask about immunotherapy eligibility, checkpoint inhibitor selection, or biomarker-guided ICI treatment decisions.

Conduct comprehensive literature research with target disambiguation, evidence grading, and structured theme extraction. Creates a detailed report with mandatory completeness checklist, biological model synthesis, and testable hypotheses. For biological targets, resolves official IDs (Ensembl/UniProt), synonyms, naming collisions, and gathers expression/pathway context before literature search. Default deliverable is a report file; for single factoid questions, uses a fast verification mode and may include an inline answer. Use when users need thorough literature reviews, target profiles, or to verify specific claims from the literature.