Brainrot

Provides persistent context management for coding sessions through FastAPI backend with SQLite storage, vector search capabilities, and semantic similarity matching to maintain architectural decisions, TODOs, and project knowledge across different development environments.

This MCP server provides persistent context management for AI coding sessions through a FastAPI backend with SQLite storage and vector search capabilities, built by ben-intrafocal using Python with SQLAlchemy, sentence-transformers, and sqlite-vec for semantic similarity search. The implementation offers project-scoped context storage with automatic project detection from working directory, supporting both basic key-value operations and advanced semantic search across stored contexts with embedding generation and similarity scoring. Features include tag-based organization, metadata support, CORS-enabled REST API, and optional vector search functionality that can be toggled via environment variables, making it valuable for developers who need to maintain coding context across different AI assistants, IDEs, and sessions while preserving architectural decisions, TODOs, and project knowledge.