scienceclaw-generation

Generate scientific hypotheses, experimental designs, and paper drafts. Use when: user asks to propose hypotheses, design experiments, or write scientific content. NOT for: data analysis or literature search.

564 stars

Best use case

scienceclaw-generation is best used when you need a repeatable AI agent workflow instead of a one-off prompt.

Generate scientific hypotheses, experimental designs, and paper drafts. Use when: user asks to propose hypotheses, design experiments, or write scientific content. NOT for: data analysis or literature search.

Teams using scienceclaw-generation should expect a more consistent output, faster repeated execution, less prompt rewriting.

When to use this skill

  • You want a reusable workflow that can be run more than once with consistent structure.

When not to use this skill

  • You only need a quick one-off answer and do not need a reusable workflow.
  • You cannot install or maintain the underlying files, dependencies, or repository context.

Installation

Claude Code / Cursor / Codex

$curl -o ~/.claude/skills/scienceclaw-generation/SKILL.md --create-dirs "https://raw.githubusercontent.com/beita6969/ScienceClaw/main/skills/scienceclaw-generation/SKILL.md"

Manual Installation

  1. Download SKILL.md from GitHub
  2. Place it in .claude/skills/scienceclaw-generation/SKILL.md inside your project
  3. Restart your AI agent — it will auto-discover the skill

How scienceclaw-generation Compares

Feature / Agentscienceclaw-generationStandard Approach
Platform SupportNot specifiedLimited / Varies
Context Awareness High Baseline
Installation ComplexityUnknownN/A

Frequently Asked Questions

What does this skill do?

Generate scientific hypotheses, experimental designs, and paper drafts. Use when: user asks to propose hypotheses, design experiments, or write scientific content. NOT for: data analysis or literature search.

Where can I find the source code?

You can find the source code on GitHub using the link provided at the top of the page.

SKILL.md Source

# Scientific Generation Skill

Generate hypotheses, experimental designs, and scientific writing across all disciplines.

## When to Use

- "Propose hypotheses for this research question"
- "Design an experiment to test..."
- "Draft a methods section for..."
- "Generate research questions for this topic"
- "Write an abstract for these findings"
- Planning new research directions

## When NOT to Use

- Running data analysis (use code-execution + scipy-analysis)
- Literature searching (use literature-search)
- Verifying claims (use scienceclaw-verification)
- Pure information extraction (use scienceclaw-ie)

## Generation Types

### 1. Hypothesis Generation
Follow the structured workflow:
1. **Observation**: State the observed phenomenon or gap
2. **Literature Context**: Reference existing knowledge and gaps
3. **Hypothesis Statement**: Formulate as testable H0/H1
4. **Variables**: Identify independent, dependent, and control variables
5. **Predictions**: State specific, measurable predictions
6. **Falsifiability**: Explain what would disprove the hypothesis
7. **Novelty Assessment**: Rate novelty (incremental/moderate/transformative)

Format: "If [independent variable] then [predicted effect on dependent variable] because [mechanism/rationale]"

### 2. Experimental Design
Include all components:
- **Objective**: Clear research question
- **Design Type**: RCT, factorial, quasi-experimental, etc.
- **Sample**: Size calculation (power analysis), selection criteria, randomization
- **Variables**: IV, DV, controls, confounds
- **Protocol**: Step-by-step procedure
- **Analysis Plan**: Statistical tests, significance thresholds
- **Ethics**: IRB/IACUC considerations
- **Reproducibility Checklist**: Materials, data sharing, pre-registration

### 3. Scientific Writing
Support all IMRaD sections:
- **Introduction**: Background, gap, objective, significance
- **Methods**: Detailed, reproducible protocol
- **Results**: Findings with statistical reporting
- **Discussion**: Interpretation, limitations, implications
- **Abstract**: Structured summary (Background, Methods, Results, Conclusions)

### 4. Research Question Generation
From a broad topic, generate:
- Descriptive questions (What/How/When)
- Comparative questions (differences between groups)
- Correlational questions (relationships between variables)
- Causal questions (cause-effect with mechanisms)

## Quality Criteria

All generated content must:
1. Be grounded in existing scientific knowledge
2. Use discipline-appropriate terminology
3. Be specific and testable (for hypotheses)
4. Include feasibility assessment
5. Consider ethical implications
6. Acknowledge limitations and assumptions
7. Cite relevant foundational work when possible

## Citation Format

When referencing prior work in generated content, use:
- Inline: (Author et al., Year) or [DOI]
- Note which citations need verification
- Distinguish confirmed vs. suggested references

Related Skills

scientific-generation

564
from beita6969/ScienceClaw

Generate scientific code, protocols, and domain-specific text with quality control

scientific-diagram-generation

564
from beita6969/ScienceClaw

No description provided.

scienceclaw-verification

564
from beita6969/ScienceClaw

Verify scientific claims, check calculations, validate experimental designs, and fact-check citations. Use when: (1) checking a claim against evidence, (2) validating statistical analyses, (3) verifying experimental reproducibility claims, (4) fact-checking references, (5) adversarial review of research. NOT for: generating new content (use scienceclaw-generation), simple QA (use scienceclaw-qa).

scienceclaw-summarization

564
from beita6969/ScienceClaw

Summarize scientific papers, datasets, experimental results, and literature reviews. Use when: (1) condensing research papers, (2) creating literature reviews, (3) summarizing experimental findings, (4) meta-analysis synthesis, (5) creating executive summaries of research. NOT for: information extraction (use scienceclaw-ie), full paper retrieval (use scienceclaw-retrieval), or writing new content (use scienceclaw-generation).

scienceclaw-retrieval

564
from beita6969/ScienceClaw

Retrieve scientific information from databases, literature, and knowledge bases. Use when: (1) finding relevant papers, (2) querying scientific databases, (3) cross-referencing findings, (4) building bibliographies, (5) systematic literature search. NOT for: answering questions (use scienceclaw-qa), summarizing (use scienceclaw-summarization), or data analysis (use code-execution skill).

scienceclaw-reasoning

564
from beita6969/ScienceClaw

Perform multi-step scientific reasoning, proof construction, causal inference, and logical argumentation. Use when: (1) deriving conclusions from premises, (2) causal analysis, (3) mathematical proofs, (4) hypothesis evaluation, (5) counterfactual reasoning. NOT for: simple factual questions (use scienceclaw-qa), data analysis (use code-execution), or literature search (use scienceclaw-retrieval).

scienceclaw-qa

564
from beita6969/ScienceClaw

Answer scientific questions across all disciplines with evidence-based responses and citations. Use when: (1) user asks factual science questions, (2) needs explanation of concepts/theories/methods, (3) multi-step scientific reasoning needed. Covers natural sciences (physics, chemistry, biology, medicine, materials, astronomy, earth science, math, CS) and social sciences (economics, sociology, psychology, political science, linguistics, history, law, philosophy, education). NOT for: opinion-based questions, non-scientific queries, or when code execution is needed (use code-execution skill).

scienceclaw-prediction

564
from beita6969/ScienceClaw

Predict scientific properties, trends, and outcomes. Use when: user asks for property prediction, trend forecasting, or model-based estimation. NOT for: historical data lookup or real-time monitoring.

scienceclaw-ie

564
from beita6969/ScienceClaw

Extract structured information from scientific texts: entities, relations, data tables, methods, results. Use when: (1) parsing papers for key data, (2) extracting experimental parameters, (3) building knowledge graphs from literature, (4) NER on scientific documents, (5) extracting methods/results sections. NOT for: summarization (use scienceclaw-summarization), full text retrieval (use scienceclaw-retrieval).

scienceclaw-discovery

564
from beita6969/ScienceClaw

Identify research gaps, synthesize cross-disciplinary insights, and generate novel hypotheses. Use when: user asks about unexplored areas, cross-field connections, or new research directions. NOT for: routine literature review or data analysis.

scienceclaw-classification

564
from beita6969/ScienceClaw

Classify scientific content by discipline, methodology, topic, and quality. Use when: user asks to categorize papers, methods, or research outputs. NOT for: simple keyword tagging or non-scientific content.

pptx-generation

564
from beita6969/ScienceClaw

No description provided.