Skills
Skills are the methods Neutropic already knows — how to search, which assumptions to check, and what a rigorous output looks like. Toggle them per project, and save your lab's own pipelines as custom skills.
Research & design
Deep investigation pipelines that plan keywords, search every database in their track, and write cited reports.
researchResearch — literature review
Investigates a topic across the paper, patent, and web tracks, then composes a single report where every sentence carries an [n] citation to a real source — uncited claims are removed, not published.
research-paperResearch — papers
The academic track. Fans out across PubMed, Crossref, OpenAlex, Semantic Scholar, Europe PMC, ERIC, and arXiv — plus Scopus, Web of Science, and KCI when connected — asks which venue tier you care about, and grounds the report in full texts rather than abstracts.
research-patentResearch — patents
Maps a technology landscape through Google Patents, PatentsView, and KIPRIS: filing-year distributions, top assignees, grant status, and milestone patents — reading actual claims before characterizing an invention.
research-webResearch — web & industry
Investigates company reports, institutional documents, and news, keeping source credibility explicit — official filings are never blended with blog posts, and every industry claim is attributed to who made it.
experiment-designExperiment design
Describe the study you want to run and it surveys how related studies actually designed theirs, then produces a structured design document with a protocol flow diagram, an expected-results chart, and a power curve.
Analysis skills
Domain-tuned analyses that run in the built-in Python kernel — each with the right checks, effect sizes, and figures for its method.
Psychometrics & stats
Classical and modern statistics with assumptions checked on every run.
- Descriptive statistics
- Reliability (Cronbach α · McDonald ω)
- Correlation analysis
- t-test (with assumptions)
- One-way ANOVA
- Exploratory Factor Analysis
- Confirmatory Factor Analysis
- Structural Equation Modeling
- Item Response Theory
- Mixed-effects models
- Regression (linear · logistic)
- Bayesian estimation (Bayes factor)
- Power analysis
Emotion engineering
Quantify affective responses and tie them to design variables.
- Semantic Differential (SD)
- Conjoint analysis
- Quantification Theory I
HCI & usability
Standard usability instruments scored and compared automatically.
- SUS (System Usability Scale)
- NASA-TLX (workload)
- UEQ (User Experience Questionnaire)
- A/B testing
Multimodal biosignals
From raw physiological signals to publication-ready figures.
- ECG / HRV (heart-rate variability)
- EDA / GSR (skin conductance)
- Eye tracking · Gaze heatmap
- EEG (band power)
- Facial expression / AU
- Voice emotion (acoustic features)
Qualitative & NLP
Make open-ended responses and interview text analyzable.
- Text overview (open-ended responses)
- Topic modeling (LDA · NMF)
- Thematic clustering (assisted coding)
- Sentiment analysis (VADER)