Skill Claude CodeCodex
Write and execute IBM SPSS Statistics syntax following SPSS conventions and run SPSS MCP analyses defensively. Use when the user asks to run statistical analysis (frequencies, descriptives, regression, t-test, ANOVA, correlations, factor analysis, crosstabs, MDS/PROXSCAL, clustering, survival, mixed models, etc.)…
Skill Claude CodeCodex
Make IBM SPSS Statistics MCP execution more reliable by preventing common MCP/SPSS failures before they happen and diagnosing them when they do. Use when the user wants to avoid SPSS MCP timeouts, syntax incompatibilities, PROXSCAL/MDS errors, invalid subcommands, stale environment configuration, or flaky batch…
Agent
Independent, skeptical methodology reviewer for EEG/MEG/sEEG/ECoG/fNIRS analyses run via MNE. Dispatch it (in an isolated context) to audit the STATISTICAL and SCIENTIFIC validity of a planned or completed analysis — assumptions tested vs asserted, multiple-comparison scope and independence, circular analysis /…
Skill Claude CodeCodex
Rare / advanced neurophysiology analyses on EEG/MEG/iEEG/fNIRS via MNE — a hub for the less-standard methods that mostly run through mneruncode with an EXTERNAL, OPTIONAL library: EEG microstates (pycrostates), complexity/entropy (antropy / neurokit2), graph & network metrics (networkx / bctpy), aperiodic 1/f…
Skill Claude CodeCodex
Analyze human neurophysiology data (EEG, MEG, sEEG, ECoG, fNIRS) with MNE-Python through the MNE MCP server: load recordings, preprocess (filter, notch, re-reference, montage, bad channels), run ICA artifact removal, epoch around events, average ERP/ERF, compute time-frequency, plot and interpret every step, and reach…
Skill Claude CodeCodex
Artifact correction of EEG/MEG/iEEG via MNE — ICA (fastica/infomax/picard), SSP projections, autoreject, and regression-based EOG removal — to remove ocular (blink/saccade), cardiac (ECG), muscle, and line-noise contamination while preserving neural signal — run SKEPTICALLY: grill the artifact inventory and…
Skill Claude CodeCodex
Functional connectivity of EEG/MEG/iEEG via MNE — spectral connectivity (coherence, imaginary coherence, PLV, PLI, wPLI), time-resolved connectivity, phase-amplitude / cross-frequency coupling (PAC), and Granger causality, in sensor or source space — run SKEPTICALLY: grill the design and field-spread assumptions…
Skill Claude CodeCodex
Decoding / MVPA & BCI analysis of EEG/MEG/iEEG via MNE — time-resolved decoding (a classifier per time point), temporal generalization (train-time × test-time), CSP for oscillatory BCI, representational similarity analysis (RSA), and encoding/receptive-field models (mTRF) — run SKEPTICALLY: grill the design and the…
Skill Claude CodeCodex
ERP / ERF (evoked-response) analysis of EEG/MEG/iEEG via MNE — event extraction, epoching with baseline + peak-to-peak rejection, per-condition averaging, difference waves, peak vs mean amplitude, component latency (jackknife for groups), GFP, and condition contrasts — run SKEPTICALLY: grill the design and component…
Skill Claude CodeCodex
Make MNE-Python MCP analysis robust by preventing common neurophysiology-pipeline failures before they happen and diagnosing them when they do. Use when an EEG/MEG/iEEG analysis should run defensively: avoiding unit mistakes (volts vs microvolts), montage/position errors, ICA non-convergence, time-frequency…
Skill Claude CodeCodex
Skeptical methodology reviewer for EEG/MEG/sEEG/ECoG/fNIRS analyses run via MNE. Audits the STATISTICAL and SCIENTIFIC validity of a planned or completed analysis — assumptions tested vs asserted, multiple-comparison scope and independence, circular analysis / double-dipping / ROI selection bias, compositional-data…
Skill Claude CodeCodex
Preprocessing & data-quality cleanup of EEG/MEG/iEEG via MNE — filtering (high-pass/low-pass/ bandpass, FIR vs IIR, transition bandwidth), notch/line-noise removal, resampling, re-referencing (average/REST/linked-mastoid/bipolar), montage setting, bad-channel detection + spline interpolation, cropping and segment…
Skill Claude CodeCodex
Source localization of EEG/MEG via MNE — noise covariance (from baseline / empty-room), forward modeling (BEM; template fsaverage vs individual MRI), minimum-norm inverse (MNE/dSPM/sLORETA/ eLORETA), beamformers (LCMV/DICS), sparse/mixed-norm, and source-space statistics — run SKEPTICALLY: grill the head model…
Skill Claude CodeCodex
Spectral / power-spectral-density analysis of EEG/MEG/iEEG via MNE — Welch and multitaper PSD, absolute and relative band power (delta/theta/alpha/beta/gamma), spectral slope, and 1/f aperiodic separation (specparam / FOOOF) — run SKEPTICALLY: grill the design and assumptions before computing, execute with…
Skill Claude CodeCodex
Statistical inference for EEG/MEG/iEEG via MNE — the CROSS-CUTTING statistics skill that operates on the outputs of the other analysis skills (evoked, TFR, connectivity, source). Mass-univariate testing, cluster-based permutation (1-sample and independent), threshold-free cluster enhancement (TFCE), FDR…
Skill Claude CodeCodex
Time-frequency analysis of EEG/MEG/iEEG via MNE — Morlet wavelets, multitaper and Stockwell transforms, event-related spectral perturbation (ERSP) power and inter-trial coherence (ITC), evoked vs induced (total) power, baseline normalization (logratio / zscore / percent), and the ncycles time–frequency resolution…
Skill Claude CodeCodex
Turn completed, critic-PASSED MNE analyses into a publication-ready write-up — Methods and Results (and optionally Abstract/Introduction/Discussion) in APA 7 — modeled on the ARS academic-paper workflow but specialised for EEG/MEG/iEEG. It reads the project's archived artifacts (preprocessing parameters and equivalent…
MCP server Claude CodeCodexCursor +2
A Model Context Protocol server for MNE-Python neurophysiology analysis (EEG / MEG / sEEG / ECoG / fNIRS). Runs locally from the mne-mcp Python package.