fmri-skill

To plan an fMRI workflow, provide a BIDS-organized dataset and request a processing plan. The skill generates a numbered execution plan for resting-state or task-based fMRI, then delegates preprocessing, denoising, ROI extraction, and connectivity steps to appropriate base tools.

The best way to coordinate reproducible fMRI pipelines is using a plan-first approach that generates an actionable execution plan without implementing concrete commands directly, ensuring clear provenance and a clean output directory layout for every processing stage.

Yes, you need BIDS-organized data for resting-state fMRI connectivity analysis. The skill requires a BIDS-structured dataset to generate a valid execution plan and delegate tasks correctly across base tool-skills for preprocessing and ROI extraction.

Yes, you can use fmriprep and xcp-d together for task-based fMRI denoising. The skill delegates processing steps to available base tools like fmriprep-tool and xcp-d, routing each stage to the correct tool skill for execution.

Executing an fMRI workflow plan requires available base tool-skills such as fmriprep-tool, hcppipeline-tool, xcp-d, and claw-shell. The skill routes and delegates specific processing steps to these tools to carry out the concrete commands.

Your fMRI workflow plan does not execute concrete commands because this skill operates strictly as a planning and delegation layer. It generates the execution plan and hands off steps to base tools like fsl-tool or nilearn to run the actual processing commands.