This repository is a starter scaffold for LifeWeaver 2.0 — an offline-first temporal-causal personal memory & action agent (MVP). It includes a Python microservice skeleton, a VS Code devcontainer, Dockerfile, and sample modules for ingestion, embeddings, KG, RAG and a safe micro-action executor.

This scaffold is intentionally minimal and designed to run in WSL Ubuntu + VS Code + Docker using the Remote - Containers extension.

• Python package skeleton: app/ with modules: ingest, embeddings, kg, rags, planner, executor
• requirements.txt with recommended packages
• Dockerfile to build a reproducible dev environment (CPU-only)
• .devcontainer/ for VS Code Remote - Containers integration
• scripts/bootstrap.sh to build and run the container locally for testing
• docker-compose.yml (optional) for future expansion
• Example CLI and FastAPI entrypoint: app/main.py

1. Build Docker image (see scripts/bootstrap.sh)
2. Open repository in VS Code and attach to the container (Remote - Containers)
3. Run the app inside the container: python -m app.main
4. Use sample endpoints / CLI to ingest files and request summaries / suggestions.

Follow the detailed setup guide in SETUP_INSTRUCTIONS.md included in this repo (step-by-step for a fresh WSL Ubuntu + VS Code + Docker setup).