ern-flankers
N2, response-related negativity, frontal theta
MNE-Python · EEGLAB · FieldTrip
Auto EEG Analysis
Reusable EEG analysis skills for coding agents.
Choose a workflow, approve the plan, and retain the executed code, results and report.
N400 in practice
N400 / ERP CORE ↗A fixed N400 tutorial, from a semantic-priming recipe to saved results from 20 ERP CORE participants.
Library
22 + 10Find a study recipe or an individual analysis skill. Open the source to inspect the inputs, methods and numerical conventions.
mmn-oddball
Auditory mismatch negativity
p300-oddball
P300 to rare targets
n170-faces
Faces versus non-face controls
n170-face-inversion
Upright versus inverted faces
n400-semantic
Semantic priming
alpha-attention-cueing
Posterior alpha modulation
resting-microstate
Resting microstate maps and dynamics
resting-spectral-connectivity
Power spectra, aperiodic fits, connectivity
complexity-anesthesia
Continuous complexity and entropy
eeg-preprocess
Filtering, reference, bad channels
eeg-ica
ICA and component labeling
eeg-epoch
Segmentation and artifact rejection
eeg-erp
ERPs and component measures
eeg-tfr
Time-frequency analysis
eeg-spectral
Power spectra and spectral parameters
eeg-stats
Statistical inference
eeg-decoding
MVPA and temporal generalization
eeg-behavior
Behavioral and trial-level associations
eeg-connectivity
Functional and directed connectivity
eeg-microstate
Maps and temporal dynamics
eeg-complexity
Entropy and complexity measures
eeg-source
Source localization
eeg-bids
BIDS-EEG conversion
eeg-qc
Data-quality checks
eeg-group-compare
Group comparisons and effect sizes
eeg-figure
Figures and data-fidelity checks
eeg-methods-text
COBIDAS-MEEG methods text
eeg-audit
Independent model review
eeg-report
Reports and reproducibility receipts
eeg-pipeline
Analysis orchestration
eeg-recipe
Recipe execution and version tracking
No matching resources. Try another term or topic.
Recipe status and validation apply to the configurations recorded in each source. Validation coverage ↗
Get started
Packages and versions ↗Use your own coding client and analysis environment.
Python / MNE
The PyPI package creates a new project with the AEA resources. Python 3.10+ and an existing MNE environment are required. The default package does not install the scientific dependencies.
In your terminal
python -m pip install auto-eeg-analysis==0.3.3
aea-eeg init ./my-eeg-study --agent codex
cd my-eeg-study
codexInside the coding client
$eeg-recipe n400-semantic --data .In your terminal
python -m pip install auto-eeg-analysis==0.3.3
aea-eeg init ./my-eeg-study --agent claude
cd my-eeg-study
claudeInside the coding client
/eeg-recipe n400-semantic --data .Place compatible data in my-eeg-study/raw/, then start your signed-in coding client in that project. Choose the recipe for your paradigm and approve the plan before computation. Configure the reviewer for the audit stage.
Install from a Git checkout
In your terminal
git clone https://github.com/dengzhe-hou/Auto-EEG-Analysis.git
cd Auto-EEG-Analysis
conda env create -f environment.yml
conda activate aeais
python tools/install_skills.py --agent codex
codexUsing MATLAB / EEGLAB?
Install AEA 0.1.3 through the EEGLAB extension manager, or use the MATLAB toolbox preview. The native MATLAB route covers explicit ERP means from existing epochs. The broader library primarily uses MNE-Python.
From a plan to saved analysis
Define the study
Choose a recipe or skills and confirm the scientific choices.
Execute the plan
The coding agent adapts and runs scripts in your environment.
Inspect and replay
Keep the executed code, numerical results, figures and review records.
What has been checked
Validation coverage ↗Numerical comparisons are tied to documented settings and saved outputs.
- ERP amplitudes
- Selected comparisons with MNE-BIDS-Pipeline, EEGLAB and FieldTrip.
- Cluster statistics
- Selected observed t maps and cluster partitions compared with FieldTrip.
- Spectral power
- Band power checked with an independent Welch implementation.
Coverage is specific to the tested output and configuration. It does not certify every generated workflow.
Build on the library
Add a recipe, improve a skill, share a tutorial, or report a first run. The contribution guide explains the expected evidence and recognition policy.
Cite the version you use
Cite AEA with the software version or commit, and cite the original methods and datasets. The Zenodo record is the frozen 6 November 2026 resource and validation archive.