About this project

PromiseLink is a personal relationship management system designed to enhance the value of business relationships. Its core functions cover event entry, entity extraction, Todo generation, promise tracking, association discovery, and data dashboards. The software adopts a Local-first architecture, where 100% of business data is stored in the user's local SQLite database. It does not provide SaaS services to ensure users have absolute ownership of their relationship assets. Architecturally, the system decouples the core algorithm layer from the LLM enhancement layer: 1. Core Algorithm Layer: Uses NetworkX, RapidFuzz, and numpy to implement entity normalization, Todo state machines, promise fulfillment tracking, association discovery, and dynamic priority scoring. It can run offline without an LLM and produces reproducible results. 2. LLM Enhancement Layer: Optional configuration of DeepSeek, OpenAI, or Anthropic APIs for entity extraction from unstructured text and response generation. The technology stack includes FastAPI (Python 3.11+), SQLAlchemy 2.0, sqlite-vec, and a Taro H5 frontend. The basic version supports manual input and CSV import, and provides one-click installation packages for macOS and Windows, allowing for quick startup without the need to configure a Docker environment.