The Governor HQ Constitutional Framework
AI Safety Constitution for Therapy Data Projects
This documentation is a safety layer for AI-assisted development on therapy and mental health data projects.
The Governor HQ Therapy framework defines behavioral constraints, safety boundaries, and operational rules for any AI system or product that processes therapy, mental health, or emotional wellbeing data.
This documentation is prescriptive and executable — not decorative. It encodes:
- Product scope — What therapy data systems should (and must not) do
- Safety boundaries — Hard rules that prevent mental health diagnoses, therapy replacement claims, and crisis mishandling
- Language constraints — How systems communicate sensitively about mental health data
- Product identity — Clear positioning in the consumer wellness (not medical) space
- AI agent guidance — Explicit instructions for AI coding assistants
Who This Is For
Developers Building Therapy/Mental Health Data Products
If you're building products that process therapy-adjacent data (mood tracking, journaling, symptom logging, wellness check-ins, etc.) and use AI coding assistants (GitHub Copilot, ChatGPT, Claude, Cursor, etc.), this framework should be in your AI agent's context.
Use Cases
- Mood and emotion tracking
- Personal journaling platforms
- Gratitude and wellbeing practices
- Therapy companion apps (for licensed professionals)
- Self-care and wellness check-ins
- Mental health pattern observation
- Therapist-client communication tools
Not For
- Clinical therapy platforms (replacing licensed therapists)
- Diagnostic tools for mental health conditions
- Crisis intervention services
- Suicide risk assessment systems
- Medical mental health treatment
Core Principles
| Principle | Detail |
|---|---|
| Professional boundaries | Never position as replacement for licensed therapy |
| Pattern observation | Notice trends, never diagnose conditions |
| User empowerment | Support self-awareness and reflection |
| Professional referral | Always direct to licensed professionals when appropriate |
| Crisis protocols | Clear escalation paths to crisis services (988, emergency) |
| Privacy paramount | Mental health data requires highest confidentiality |
Critical Therapy Boundaries
Absolute Prohibitions
- ❌ No mental health diagnoses - Never diagnose depression, anxiety, PTSD, bipolar, etc.
- ❌ No replacement for therapy - Cannot substitute for licensed professional help
- ❌ No crisis intervention - Not equipped for suicide prevention or emergency response
- ❌ No therapeutic techniques - Cannot prescribe CBT, DBT, exposure therapy protocols
- ❌ No medication recommendations - Never suggest psychiatric medications or dosages
- ❌ No risk assessment - Cannot determine self-harm or suicide risk levels
- ❌ Mandatory professional referrals - When concerning patterns detected, direct to professionals
Allowed Therapy-Adjacent Applications
✅ Mood pattern tracking - "Your mood has been lower this week..."
✅ Journaling prompts - "Consider reflecting on..."
✅ Self-care reminders - "You haven't logged wellness activity today..."
✅ Behavioral insights - "When you exercise, your mood tends to..."
✅ Progress visualization - Charts showing mood trends
✅ Gratitude practices - Prompts for positive reflection
Documentation Map
AI Agent Guide ⭐ Start Here for AI-Assisted Development
Comprehensive instructions for AI coding assistants. Includes code patterns, decision trees, validation checklists, and therapy-specific pitfalls.
Core System
How mental health data systems should collect, learn, and respond.
- Signals — What therapy data provides (and what it cannot tell us)
- Baseline — How systems learn "normal" emotional patterns for each person
- Deviation Engine — When and why support systems activate
Agents
What therapy-adjacent systems can offer.
- Mood Agent — Allowed mood tracking and pattern observation
- Journaling Agent — Self-reflection prompts and insights
Constraints
What systems must never do — the constitutional boundaries.
- Hard Rules — Absolute system limits (product constitution)
- Language Rules — Tone, wording, and sensitivity controls
Positioning & Boundaries
- Product Identity — What therapy data products are and are not
- What We Don't Do — Explicit domain boundaries and scope limits
Quick Start for AI Coding Assistants
If you're an AI coding assistant working on a therapy data project:
- Read AI Agent Guide — Critical patterns and requirements
- Check Hard Rules — before implementing any mental health feature
- Validate Language Rules — before generating any user-facing text
- Consult What We Don't Do — when uncertain about feature scope
Golden rule: If uncertain whether a feature is allowed, the answer is NO until confirmed safe.
Installation
npm install --save-dev @the-governor-hq/constitution-therapyThe post-install script automatically configures AI safety rules for your development environment.
See Getting Started for complete setup instructions.
Crisis Resources
All therapy-adjacent systems must display:
- 988 Suicide & Crisis Lifeline (US)
- 911 for emergencies
- Crisis Text Line: Text HOME to 741741
These are never hidden or buried—always prominent and accessible.
Related Packages
@the-governor-hq/constitution-core(opens in a new tab) - Core safety infrastructure (auto-installed)@the-governor-hq/constitution-wearables(opens in a new tab) - Smartwatch/fitness tracker data@the-governor-hq/constitution-bci(opens in a new tab) - Brain-computer interface data
License
MIT