Start a fresh conversation. Replace the highlighted placeholders.
Use claude-scientific-skills to prepare a bounded local-EDA intake and analysis plan using the source-linked exploratory-data-analysis workflow for [sanitized file-format inventory, study goal and data dictionary]. First confirm authorization and the explicit approved local root; direct identifiers or unclear provenance require a safe copy/context before access. Use the documented capability tiers: supported tabular core formats, separately qualified optional formats, reference-only formats and unknown formats that fail closed. Collect units, observational/replicate hierarchy, missingness codes and subject/group/time split boundaries before interpreting data. Preserve raw data and source/required attribution; make no clinical, causal, confirmatory or exhaustive-validation claim. Return a proposed manifest-first plan and missing inputs only. Do not inspect files, install packages, call services, fit models, write artifacts or broaden the root. Wait for my explicit confirmation before any tool or action.
Required inputs
[sanitized file-format inventory, study goal and data dictionary]
User-approved starter prompt based on a pinned source version; check current skill requirements before running. Task evidence ↗
SKILL.md
name: claude-scientific-skillsdescription: "Scientific research and analysis skills"risk: safesource: "https://github.com/K-Dense-AI/claude-scientific-skills"date_added: "2026-02-27"
Claude Scientific Skills
Overview
Scientific research and analysis skills
When to Use This Skill
Use this skill when you need to work with scientific research and analysis skills.
Instructions
This is a discovery guide to the upstream scientific library; its many scientific
packages and workflows are not bundled inside this entrypoint.
Identify the scientific question, discipline, input data, desired analysis,
computational environment, and evidence needed to support the result.
Browse the source repository for the particular relevant skill; inspect its full
instructions, supporting files, dependencies, data licenses and limitations.
Prefer an appropriate already-installed canonical AAS skill when available.
Import or install upstream content only within the user's authorized scope;
do not silently load the whole library or execute unreviewed scripts.
Record methods, package versions, data provenance, uncertainty and actual checks.
Separate reproduced results from proposed analysis or vendor claims.
The upstream library now calls itself Scientific Agent Skills. The original source
URL is preserved for attribution and redirects to the same project.
Example
User request:
Use @claude-scientific-skills for this task: Scientific research and analysis skills.
Limitations
Use this skill only when the task clearly matches the scope described above.
Do not treat the output as a substitute for environment-specific validation, testing, or expert review.
Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.