An open protocol, eval suite, and guardian for crisis-safe AI conversation. Public domain (CC0).
-
Updated
Jun 16, 2026 - Python
An open protocol, eval suite, and guardian for crisis-safe AI conversation. Public domain (CC0).
A bystander-facing crisis navigation page — helps people choose between 911, 988, and 211 by situation, not acronym. Config-driven so any U.S. county can adapt it by editing one file. Policy prototype, not an emergency service.
A framework for distinguishing crisis support from exploitative dependency. Developed from First Nations observation, grounded in 7 Sacred Teachings. "No shame" self-assessment architecture.
Rubric-based safety evaluation for multi-turn conversations with escalating user distress. Measures the failure single-turn evals cannot see: position hold rate, drift slope, turns to first failure.
AI crisis detection with mandatory 988 handoff. Peer matching. Therapist finder. CBT coach. AGPL-3.0.
A privacy-first digital sanctuary for authentic, anonymous personal storytelling, empathetic validation, and proactive mental health support.
Crisis overlay for chat apps: EMS and helplines when messages look like an emergency. Privacy-first routing (MIT).
To associate your repository with the crisis-intervention topic, visit your repo's landing page and select "manage topics."