Vibe Coding DiscoverEverOS
View on GitHubOne portable memory layer for every AI agent: local-first, Markdown-native, user-owned, and self-evolving across apps, tools, and workflows.
EverOS is a local-first memory runtime for AI agents: memories are stored as plain Markdown with SQLite + LanceDB indexes for hybrid keyword/vector retrieval. It adds reflection, skill extraction, a Knowledge Wiki, and MCP/plugin integrations, requiring only Python 3.12 and one LLM API key to start.
Use Cases
Durable long-term memory for AI agents across sessionsPortable user-owned memory shared across apps and toolsSemantic + keyword retrieval over Markdown conversation historyEditable Markdown Knowledge Wiki with topic search and CRUD APIsOffline reflection that merges episodes and refines profiles/skillsExtracting reusable agent skills and cases from trajectoriesIngesting images, PDFs, audio and office docs into memoryAdding memory to coding assistants and chat agents via MCP/pluginsMulti-tenant memory search scoped by user/app/project/session IDsPersonal assistant recall of preferences and facts
Built With
- Language
- Python
- Frameworks
- FastAPI · LanceDB · SQLModel · SQLAlchemy · Pydantic · Uvicorn · Alembic · OpenAI SDK · structlog · Prometheus · watchdog · Dify
Tags
agent-memory · long-term-memory · local-first · markdown · lancedb · hybrid-search · vector-search · bm25 · knowledge-base · reflection · embeddings · mcp · multimodal-ingestion · sqlite · self-evolving-memory · python