GoKawiilEngrim launches as local-first SQLite memory layer for AI coding agents
A new open-source tool called Engrim gives AI coding assistants persistent, project-level memory stored in a local SQLite database rather than tied to any single vendor's cloud. It lets developers switch between tools like Claude Code, Cursor, Windsurf and Google Antigravity mid-project while retaining architectural decisions and constraints, using a hybrid search combining SQLite FTS5 keyword search with lightweight vector embeddings. The creator reports it was validated over 105 sessions on a 50,000-line trading codebase with zero regressions and no loss of context across model switches.
GoKawiil's interpretation of the reporting above, not reported fact.
AI coding assistants currently trap project context inside proprietary sessions, forcing developers to either stick with one vendor or re-explain decisions every time they switch tools or clear a session. A portable, local memory layer addresses both the rising cost of huge context windows and vendor lock-in, letting developers treat models as interchangeable while keeping project knowledge independent and under their own control.
- Engrim stores project memory locally in SQLite, decoupling it from any single AI model or vendor.
- It combines keyword search (FTS5/bm25) with vector embeddings to quickly reload only relevant context.
- A 105-session case study on a large trading codebase reported zero regressions and no context loss across model switches.
O'Reilly "Using SQLite" by Jay A. Kreibich — If a local-first SQLite memory engine like engrim has you curious about what's under the hood, this O'Reilly guide walks through SQLite's design, SQL usage, and embedding it in your own applications. It's a solid desk reference for developers building project-scoped local databases instead of leaning on cloud silos.
See O'Reilly "Using SQLite" by Jay A. Kreibich on Amazon → Affiliate link — we may earn a commission on purchases, at no extra cost to you. Product picked by AI based on this article; it is not a tested recommendation.Source: github.com, 2026-09-07
Published there as: “Show HN: Engrim – A universal, local-first SQLite memory engine for AI CLIs”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.