BIDS Neuroimaging Data Standard Assistant

What it does

BIDS (Brain Imaging Data Structure) is the community standard for organizing neuroscience and biomedical research data with consistent directories, filenames, and metadata. This skill guides Claude through:

• Laying out all 11 modalities: MRI (anat, func, dwi, fmap, ASL), PET, microscopy, EEG/MEG/iEEG/EMG, NIRS, motion capture, and behavioral data
• Writing dataset_description.json, participants.tsv, events.tsv, and scans.tsv
• Querying datasets with PyBIDS BIDSLayout by subject/session/task/run, including SQLite index caching
• Checking compliance with the Deno-based bids-validator and configuring .bidsignore
• Converting DICOM to BIDS via HeuDiConv, dcm2bids, or BIDScoin
• Running BIDS-Apps (fMRIPrep, MRIQC, QSIPrep) and producing valid derivatives

It also ships eight diagnosed failure modes with fixes (missing dataset_description, absent SliceTiming, TSV encoding problems, phase-encoding axis confusion, slow indexing) plus best practices such as defacing, metadata inheritance, BIDS URIs for provenance, and DataLad versioning. Bundled references include the machine-readable BIDS schema, the current BEP table, per-modality metadata fields, and conversion-tool recipes.

Who it's for

• Researchers preparing datasets for OpenNeuro, DANDI, or journal data-sharing requirements
• Imaging analysts who need reproducible DICOM-to-BIDS pipelines
• Data managers harmonizing datasets across multiple labs
• Graduate students still learning BIDS filename grammar and required metadata

Examples

1. DICOM conversion: "Write a HeuDiConv heuristic for this scanner folder" → Claude follows the reconnaissance → heuristic → convert sequence and can apply ReproIn naming.
2. Validation debugging: "The validator says 'Not a BIDS dataset' — why?" → it identifies the missing root dataset_description.json and generates a minimal version.
3. Query script: "List paths and RepetitionTime for every task-rest bold file" → it produces PyBIDS code with entity filters and database_path caching for speed.

Install the skill I found on Claude Skill Mart.
Copy the skills/bids folder from the GitHub repo K-Dense-AI/scientific-agent-skills into my ~/.claude/skills/bids/.
When it's done, tell me in one line what this skill can do.