Tag
Release of github-to-sqlite version 2.9.1, a tool for exporting GitHub data to SQLite databases, posted by Simon Willison.
The article introduces 'memoryfields', a portable file format for AI agent memory that uses Markdown files and optional SQLite indexes, advocating for a simpler, data-driven approach over existing complex systems.
OpenRun introduces built-in Litestream support for SQLite apps, enabling automatic replication to object storage and seamless deployment on Docker and Kubernetes with GitOps integration.
Squad is a command-line tool that enables multiple AI agents to collaborate in a terminal using shell commands and SQLite for real-time communication, without background processes.
The article introduces the concept of executables as SQLite databases, demonstrating a proof-of-concept webserver that stores its program, routes, and state in a single file.
The article discusses a long-standing bug in SQLite's WAL-reset mechanism that caused data loss and corruption, reproducing it with a 100-line C workload to demonstrate the race condition.
This article explains how to use SQLite's JSON support and generated columns to treat SQLite as a document database, enabling data extraction and indexing from JSON columns.
A blog post explains a Linux pattern where a SQLite database file is formatted to act as an ELF executable, using a custom interpreter and binfmt_misc for direct execution.
The article explores replacing the ELF executable format with SQLite, introducing a prototype called SELF that enables executables to be SQLite databases and discusses the benefits and technical implications.
The author replaced a file-based agent handoff system with a self-hosted message board using SQLite, tested it in a real task, and documented failure points.
Hillock v0.5.0 is a lightweight neuro-symbolic memory engine for AI agents that manages persistent memory efficiently using structured triples in SQLite and hypervectors, designed for local setups with limited VRAM.
The author describes building a memory layer for news-reading AI agents to handle deduplication and improve continuity across sessions, using SQLite and vector storage with a Python API, CLI, and MCP server.
The article advocates for SQLite as a versatile and stable database solution, highlighting its ability to replace multiple other tools in various tech stacks.
A developer created an open-source local memory layer called Memmy to share context between Claude Code and Cursor, eliminating the need to re-explain architecture when switching tools.
BriskDB is an alpha-stage database tool that transforms SQLite files into a sharded system with parallel writes, PostgreSQL compatibility, and embedded Rust/Python APIs, aiming to scale SQLite while maintaining its simplicity.
TraceMotive v0.2.0, a local AI-agent debugging tool, has been released with improvements like persistent storage, one-command startup, and trace comparison features, addressing feedback from the initial version.
sqlite-utils 4.2.1 fixes a crashing bug caused by an unlisted typing-extensions dependency, and adds a smoke test to verify CLI works without dev dependencies.
sqlite-utils 4.2 improves the transform() feature with support for complex schema definitions, new introspection properties, and other smaller changes, with contributions from multiple developers.
MCP-Memory is an open-source MCP server that gives AI agents persistent memory using Google's Open Knowledge Format and SQLite FTS5 for fast retrieval and search.
This article summarizes the key points of Richard Hipp's 2024 talk on SQLite design principles and philosophy, explaining SQLite's embedded, single-file, low-complexity features and how it works.