Обнаружение кризисных сигналов и эскалация к специалисту

★ 6.8 · general

crisis-detection-intervention-ai is a Claude Code skill that detects crisis signals in user-generated text using specialized NLP models — MentalBERT (89% accuracy), GPT-4 with few-shot prompting (92%), and RoBERTa-Mental — combined with safe intervention protocols. It identifies suicidal ideation, self-harm mentions, and substance relapse risk, then flags content for human counselor review and immediately surfaces crisis resources such as the 988 Suicide & Crisis Lifeline and Crisis Text Line (text HELLO to 741741). The skill uses Read, Write, Edit, and Bash (npm) tools and follows a detection-flag-notify flow that deliberately avoids automated AI responses to distressed users. Built for mental health journaling apps, recovery communities, and online therapy platforms where suicide prevention NLP and ethical escalation pipelines are required. It is not a substitute for licensed therapists, emergency services, or medical diagnosis.