Atomic Memory
Memory infrastructure for AI agents and applications
Visit docs.atomicstrata.ai ↗What is Atomic Memory?
Atomic Memory is an open-source semantic memory system designed for AI agents and applications. It provides durable context across sessions without locking you into a specific model, framework, or deployment method. Users can start with managed Hosted Cloud, run the Core locally, or integrate via TypeScript SDK and MCP server. The system emphasizes inspectable, correction-aware memory that agents can audit and verify.
The platform offers a standardized architecture with five independently replaceable domains: composition, stores, scope, SDK, and API. It includes a TypeScript client for memory and artifact storage, a language-agnostic HTTP API, and integrations with Claude, Vercel AI SDK, and LangChain. Setup requires only `am init` with no Docker needed, and the open-source Core can run locally for free.
Momentum
Daily activity over the last 30 days: upvotes, unique visits, new GitHub stars and Hacker News points.
How do we know it's vibe coded?
vibetrend discovered Atomic Memory automatically. Evidence: GitHub atomicstrata/atomicmemory; repo has AGENTS.md, CLAUDE.md, GEMINI.md. View source.
Similar vibe-coded apps
Open source AI gateway routing across 350+ providers through one endpoint
Open-source AI job search agent that runs locally in your CLI
A second brain for the AI era. Markdown notes with Git and AI agents.
Local code graph for AI-assisted reviews with 63x context reduction
AI-native engineering philosophy and plugin for coding agents
Open-source framework for building autonomous agents with purpose-built UIs
Lightweight personal AI agent you run locally, with your keys or local models
Sovereign agent operating system for enterprises to own their AI knowledge