Show HN: Engrim – A universal, local-first SQLite memory engine for AI CLIs

Engrim is a local-first SQLite memory engine that stores curated episodic working memory for AI CLIs, enabling model switching (Gemini 3.8, Claude 3.7 Sonnet, GPT-4o) without losing project state. It uses SQLite FTS5 with bm25 search and static vector embeddings for zero-latency hybrid retrieval, cutting reloaded context cost by >99% in production algorithmic-trading tests. Adapters and hooks support Google Antigravity, Claude Code, and Cursor MCP, auto-detecting and configuring installed environments.