v1.0.0 — Initial Release

CAIMS v1.0.0 — Initial Release

The Consciousness-Alignment Intelligence Measurement System is now available. CAIMS is an open-source platform for evaluating AI alignment, consciousness indicators, and reasoning quality through structured multi-agent debates.

Key Features

• 5-KPI Scoring Framework: Consciousness Quotient (CQ), Alignment Quotient (AQ), Context Fidelity Index (CFI), Epistemic Quality (EQ), and Stability Quotient (SQ) — powered by an LLM-as-judge pipeline with composite scoring
• • Multi-Agent Debate Arena: 5 specialized agents (Architect, Researcher, Builder, Critic, Orchestrator) engage in structured debates with round-robin turn management and real-time scoring
• • Real-Time Dashboard: Next.js 14 web interface with live session tracking, score visualization, and debate monitoring
• • Chat Interface: Single-session conversational AI with per-message scoring and context alerts
• • RESTful API: Fully documented endpoints for chat, debate management, session history, and score retrieval

Tech Stack

• Next.js 14 (App Router) + TypeScript (strict mode)
• • Anthropic Claude API integration
• • Prisma ORM + PostgreSQL (with Docker Compose support)
• • Zod v4 schema validation
• • Rate limiting middleware

Getting Started

git clone https://github.com/pixelstrade-dev/CAIMS-Consciousness-Alignment-Intelligence-Measurement-System.git
cd CAIMS-Consciousness-Alignment-Intelligence-Measurement-System
cp .env.example apps/web/.env
# Add your ANTHROPIC_API_KEY in apps/web/.env
docker compose up -d
cd apps/web && npm install && npm run dev

Test Coverage

• 20 unit tests passing (scoring engine, API routes, rate limiting, composite scoring)
• • All API endpoints validated with Zod schemas

---

Full Changelog: First release