@chen-feng123

judgment-enhancement-engine applies monte carlo tree search with risk-adjusted utility scoring to help agents evaluate multi-step action sequences under uncertainty. configurable lookahead depth, breadth, and risk tolerance allow tuning for different decision environments.

cognitive-enhancement-engine provides working memory, tf-idf vector search, goal planning, reasoning and reflection for python agents. no external dependencies. lacks explicit failure modes and recovery procedures.

ai-diabetes-coach provides blood glucose logging, risk stratification, personalized diet and exercise guidance, and insulin dose estimation aligned to ada 2025 standards with emergency value interception and audit logging.

health-reasoner scores daily habits (sleep, diet, exercise, stress, substance use) against rule-based thresholds and suggests improvements. pure python, no dependencies, no medical claims.

Persistent memory engine for AI agents with semantic recall, hotness prioritization, importance weighting, time decay, and auto-compaction. Zero external dep...

AI 心脏康复管理系统 — 基于 ACC/AHA 指南的安全增强型康复管理 Web 应用。症状预警、风险分层、个体化运动处方、药物依从性提醒。