Документация для систем поддержки клинических решений

★ 7.0 · research

clinical-decision-support is a Claude Code skill that helps prepare and validate research-only documentation for clinical decision-support evaluation: intended-use artifacts, evidence profiles, survival analysis plans, aggregate biomarker and model evaluations, de-identification checklists, and decision-logic traceability matrices. It ships with 7 JSON templates and 7 Python 3.11+ scripts built on the standard library — no external dependencies, network calls, API keys, or remote services; all source files stay local. Pharmaceutical research teams and clinical investigators use it to produce GRADE evidence-profile checklists, aggregate cohort tables with disclosure controls, survival-analysis plan reviews, and governance documentation for AI/ML models. A hard safety boundary prohibits any patient-specific use: the skill must never diagnose, recommend treatment, calculate doses, or support bedside or live clinical operation — every output remains a draft until reviewed and approved by qualified human experts.