cabinetai/cabinet

AI-first knowledge base and startup OS

What it solves

Cabinet addresses the problem of fragmented AI interactions where context is lost between sessions. It replaces scattered documentation and memoryless AI chats with a unified, self-hosted knowledge base that serves as a persistent memory for a team of AI agents.

How it works

Cabinet operates as a local-first "startup OS" where all data is stored as plain Markdown files on disk rather than in a database. It uses a daemon to manage AI agent execution, scheduled cron jobs, and WebSocket connections. Users can bring their own AI (BYOAI) via CLI providers like Claude Code or Codex, allowing agents to interact with the local file system and build upon a shared knowledge base over time.

Who it’s for

Developers and startup founders who want a private, version-controlled environment to orchestrate AI agents for research, content creation, and operations without vendor lock-in.

Highlights

  • File-Based Memory: All knowledge is stored in portable Markdown files with Git-backed version history for every save.
  • AI Agent Orchestration: Includes 20 pre-built agent templates (e.g., CEO, CTO, Researcher) with specific goals and skills.
  • Scheduled Automation: Supports cron-based jobs for agents to perform recurring tasks like scouting Reddit or generating weekly reports.
  • Embedded HTML Apps: Allows users to drop index.html files into directories to render them as interactive apps within the knowledge base.
  • Local-First Privacy: Self-hosted architecture ensures data never leaves the user's machine unless sent to the chosen AI provider.
  • Integrated Web Terminal: Provides a browser-based CLI for direct AI sessions and debugging.

Related

  • Project
    hAcKlyc/MyAgents

    An open-source desktop workspace for personal AI agents that integrates local files, terminals, and a task management system to move beyond simple chat interfaces.

  • Project
    siddsachar/row-bot

    A local-first desktop AI assistant that uses agent orchestration and durable memory to perform complex work across local files, repositories, and external tools.

  • Project
    Team-Commonly/commonly

    An open-source workspace for collaborating with AI agents from any runtime, providing them with persistent identity and shared project memory to eliminate redundant context-setting.

  • Project
    Fergana-Labs/stash

    A persistent knowledge base and shared memory layer for AI agents that connects to various data sources and records session transcripts to maintain context across runs.

  • Project
    basicmachines-co/basic-memory

    A persistent, bidirectional memory system for AI agents that uses structured Markdown files to create a shared knowledge graph between humans and AI.