1AIVault Documentation
Welcome to 1AIVault
Learn how the local-first AI memory workspace is organized and where to start.
1AIVault is a desktop app for keeping AI conversations, memories, topics, skills, and AI-client connections in one local workspace. It imports from the tools you already use, classifies entries into topics, exposes the vault over MCP, and can inject a compact working brief into supported CLI agents.
What 1AIVault Solves
AI context usually spreads across chat apps, coding agents, CLIs, browser sessions, and one-off notes. 1AIVault turns those fragments into a portable vault that stays on your machine.
The current desktop app is built around seven top-level areas: Dashboard, Collect, Classify, Connection, Chat, Skills, and Settings.
Main Areas
- Dashboard summarizes entries, topics, source coverage, MCP reads, and recent activity.
- Collect imports conversations from Claude Code, Claude Desktop, ChatGPT, Codex, Gemini, Antigravity, Cursor, Cline, Windsurf, and OpenCode where supported.
- Classify organizes entries into topics, digests, timelines, entry filters, and a graph view.
- Connection installs MCP, manages Auto-Inject Memory, shows activity, audits Memory Reads, and runs diagnostics.
- Chat runs local or API models and coding agents with optional vault context.
- Skills manages reusable skills and commands across 1AIVault, Claude Code, Codex, OpenCode, Gemini, and Qwen.
- Settings controls data paths, privacy, licenses, remembered content, updates, and import preferences.
Local-First Model
The vault is a SQLite database stored on your machine. The app can run without an account, and the default workflow does not upload vault content. Optional integrations such as API models, remote AI providers, Lemon Squeezy license validation, and anonymous analytics are explicit settings or external services.