DigniFi - Trauma-Informed Bankruptcy Filing Platform

Mission: Democratize access to bankruptcy relief for low-income, pro se Americans by translating complex legal processes into plain-language guidance and auto-populated court forms.

CRITICAL: This platform provides legal INFORMATION, never legal ADVICE. All development must respect UPL (Unauthorized Practice of Law) boundaries.

Project Status

MVP Complete — Full-stack application validated through AI persona testing (Mar 2026)

Pilot District: Northern District of Illinois (ILND) MVP Scope: Chapter 7 bankruptcy filing assistance Next Milestone: Paper prototype testing with target demographic

Test Results (Mar 2026)

Suite	Count	Status
Backend (pytest)	413	Passing
Frontend (vitest)	165	Passing
E2E persona tests	5/5 personas	All steps green

Quick Start

Prerequisites

• Docker & Docker Compose (or Colima on macOS)
• Python 3.11+ (for local development without Docker)
• Node.js 18+ (for local development without Docker)

Running with Docker (Recommended)

# 1. Copy environment file
cp backend/.env.example backend/.env

# 2. Generate encryption key
python3 -c "from cryptography.fernet import Fernet; print(Fernet.generate_key().decode())"
# Add the output to FIELD_ENCRYPTION_KEY in backend/.env

# 3. Start services
docker compose up -d

# 4. Run migrations
docker compose exec backend python manage.py migrate

# 5. Load ILND district data
docker compose exec backend python manage.py loaddata ilnd_2025_data

# 6. Seed demo personas (optional — for testing)
docker compose exec backend python manage.py seed_demo_data

# 7. Access the application
# - Frontend: http://localhost:5173
# - Backend API: http://localhost:8000/api
# - Admin: http://localhost:8000/admin

Verify Everything Works

# Backend tests
docker compose exec backend pytest -q

# Frontend tests
cd frontend && npx vitest run

# Quick persona smoke test (requires Playwright)
python3 test_maria_quick.py

Code Quality Setup (Recommended)

Set up automatic linting and formatting before every commit:

# One-time setup
./scripts/setup-linting.sh

# Or manually:
pip install pre-commit
pre-commit install

This installs pre-commit hooks that automatically:

• Format code with Prettier (frontend) and Black (backend)
• Lint with ESLint (frontend) and Ruff (backend)
• Fix trailing whitespace and other issues

See docs/LINTING_SETUP.md for details.

Local Development (Without Docker)

Backend:

cd backend
python3 -m venv venv
source venv/bin/activate
pip install -r requirements/development.txt
cp .env.example .env
# Edit .env with your local database settings
python manage.py migrate
python manage.py runserver

Frontend:

cd frontend
npm install
npm run dev

Project Structure

dignifi/
├── backend/                 # Django REST API
│   ├── apps/
│   │   ├── users/          # Authentication & profiles
│   │   ├── audit/          # UPL compliance + analytics logging
│   │   ├── districts/      # District-specific rules & data
│   │   ├── intake/         # Data collection (6-step wizard)
│   │   ├── eligibility/    # Means test calculator
│   │   └── forms/          # 13 bankruptcy form generators
│   ├── config/             # Django settings (base, dev, prod)
│   └── requirements/       # Python dependencies
├── frontend/               # React 19 + TypeScript SPA
│   ├── src/
│   │   ├── components/     # UI components (wizard, forms, compliance, survey)
│   │   ├── pages/          # IntakeWizard, FormDashboard, Login, Register
│   │   ├── context/        # AuthContext, IntakeContext
│   │   ├── api/            # Typed API client
│   │   └── utils/          # Analytics, helpers
│   └── e2e/                # Playwright page objects & journey specs
├── docs/                   # Reference documentation
│   ├── testing/            # Persona briefs & orchestration protocol
│   ├── reports/            # Usability test reports
│   └── plans/              # Design documents
├── Product Docs/           # PRD, briefs, architecture analysis
├── test_persona_full_flow.py  # 5-persona E2E test script
└── test_maria_quick.py        # Quick smoke test

Technology Stack

• Backend: Python 3.11, Django 5.0, Django REST Framework, PostgreSQL 15
• Frontend: React 19, Vite 7, TypeScript (Context API for state management)
• PDF Generation: PyPDF2 (13 form generators)
• Authentication: JWT (djangorestframework-simplejwt) with silent refresh
• Encryption: django-encrypted-model-fields (Fernet) for PII
• Testing: pytest, vitest, Playwright, vitest-axe (accessibility)
• CI/CD: GitHub Actions (lint, backend tests, frontend tests, E2E)
• Infrastructure: Docker, Docker Compose

Core Features

• User authentication with JWT (access in memory, refresh in localStorage)
• 6-step intake wizard: Debtor Info → Income → Expenses → Assets → Debts → Review
• Chapter 7 means test calculator (11 U.S.C. § 707(b))
• Fee waiver eligibility assessment (28 U.S.C. § 1930(f))
• 13 bankruptcy form generators (Form 101 through Schedules A/B–J)
• UPL confirmation modal gates all form generation
• Post-task usability survey (3 Likert + 2 open text)
• Fire-and-forget analytics via AuditLog API
• WCAG 2.1 AA accessibility (ARIA labels, keyboard nav, skip links)
• Trauma-informed language throughout ("amounts owed" not "debt")

Development Guidelines

UPL Compliance (CRITICAL)

Never use these phrases:

• "You should file Chapter 7"
• "I recommend..."
• "Based on your situation, file X"

Always use information-only language:

• "You may be eligible for Chapter 7 if..."
• "Chapter 7 typically requires..."
• "This information is not legal advice"

See docs/UPL_COMPLIANCE.md for complete guidelines.

Plain Language

• Target 6th-8th grade reading level (Flesch-Kincaid)
• Define legal terms on first use
• Use trauma-informed language (see docs/TRAUMA_INFORMED_DESIGN.md)

Data Security

• Encrypt all PII (SSN, income data, creditor information)
• Never log PII in error messages
• Use HTTPS in production
• JWT access tokens expire after 30 minutes
• Synthetic SSNs for testing use IRS-reserved 900-xx range

Testing

# Backend (413 tests)
docker compose exec backend pytest -q

# Frontend (165 tests)
cd frontend && npx vitest run

# Full persona E2E (5 personas × full flow)
python3 test_persona_full_flow.py

# Quick smoke test (Maria only)
python3 test_maria_quick.py

API Documentation

Key endpoints:

Method	Path	Description
POST	/api/token/obtain/	JWT login
POST	/api/token/refresh/	Refresh access token
POST	/api/intake/sessions/	Create intake session
PATCH	/api/intake/sessions/{id}/	Update session data (nested serializer)
POST	/api/intake/sessions/{id}/complete/	Finalize intake
POST	/api/intake/sessions/{id}/calculate_means_test/	Run means test
POST	/api/forms/generate_all/	Generate all 13 forms
GET	/api/forms/?session={id}	List generated forms

Contributing

1. Review docs/UPL_COMPLIANCE.md — UPL boundaries are non-negotiable
2. Follow plain-language guidelines (docs/PLAIN_LANGUAGE_GUIDE.md)
3. Use trauma-informed design principles (docs/TRAUMA_INFORMED_DESIGN.md)
4. All PRs must pass CI (lint, backend tests, frontend tests)
5. Maintain 80%+ test coverage

Deployment

Production checklist:

• Generate new DJANGO_SECRET_KEY
• Generate new FIELD_ENCRYPTION_KEY
• Set DEBUG=False
• Configure HTTPS/SSL
• Set up database backups
• Configure email service
• Review security settings in config/settings/production.py

License

[To be determined - likely AGPL-3.0 or similar copyleft license]

Contact

Project Lead: Courtney Richardson Organization: DigniFi (Social Impact Startup)

---

Legal Disclaimer: This software provides legal information only and does not constitute legal advice. Users should consult a licensed attorney for legal advice specific to their situation.