Release v1.0.0

What's New in v1.0.0

Core Features

• Dual-Phase Architecture: Vigilia (wakefulness) for fast responses, REM phase for autonomous learning
• Fractal Graph: RAPTOR recursive structure with semantic clustering
• Storage Engine: SurrealDB with SSD persistence and HNSW indices for vectors
• Multi-User Support: Strict namespace separation (global/personal) with per-user scopes
• Optimized Navigation: O(m log^(2/3) n) using hopsets for shortest paths

API Endpoints

• POST /v1/ingest: Ingest files (PDF, images with OCR, text)
• POST /v1/remember: Store episodic/conversational memory
• POST /v1/ask: Query fractal graph for LLM context retrieval
• POST /v1/sync_rem: Trigger REM phase (consolidation + web learning)
• PATCH /v1/memory: Manual knowledge updates/corrections

Technical Highlights

• Local embeddings via FastEmbed (no external API needed)
• Local web search via SearXNG integration
• Persistent SSD storage
• Docker Compose for easy deployment

Roadmap

• Fractal data model
• SurrealDB with HNSW
• REST API (Axum)
• RAPTOR clustering
• SSSP navigation with hopsets
• Web UI (React/Tauri)

Installation

Quick Start

git clone <repo-url>
cd fractalmind
./dev.sh run

Docker Services

• SurrealDB (port 8000)
• SearXNG (port 8080)
• FractalMind API (port 9000)

See README for full configuration options.