About this project
Octop is an open-source, self-hosted AI assistant platform designed for households and small teams. It runs as a single process that serves a web dashboard, CLI, IM channels (Feishu, DingTalk, QQ, Telegram, WeCom, Discord, and more), and cron automation, all sharing one control-plane database under `~/.octop/` (SQLite by default, PostgreSQL optional).
Key features include:
- **Multi-user expert team**: One admin, shared household; built-in expert library with switchable specialists per scenario.
- **MBTI personas**: 16 personality templates plus an interactive quiz for distinct agent characters.
- **Security built-in**: JWT multi-user isolation, tool approval, shell command guardrails, and PII redaction.
- **Connector ecosystem**: Tencent suite (Docs, Weibo trends, News), OAuth and MCP gateway.
- **Pluggable backends**: Local disk, Docker containers, PostgreSQL, or COS/S3.
- **Portable memory**: Powered by harness-memory; memory migrates with the workspace.
- **Knowledge base**: RAG over documents for semantic retrieval.
- **Plugins**: Third-party plugin support with bundled plugins.
- **ACP bidirectional**: `octop acp` for IDE/terminal AI; delegate to OpenCode / Claude Code with permission gates.
- **Terminal AI+**: Interactive shell in the browser with AI-assisted command execution.
- **Browser AI+**: Headless Chromium sessions for web automation, screenshots, and remote browsing.
- **Remote desktop**: Live screen and input from the dashboard on Linux, Windows, and macOS.
- **Self-hosted**: All data under `~/.octop/`.
Core technology: Python 3.12+, FastAPI + uvicorn, harness-agent runtime, harness-gateway for IM, harness-memory, harness-browser (CDP-based), React 18 + TypeScript + Vite + Ant Design frontend, APScheduler for scheduling, and agent-client-protocol for ACP.
Quick start: One-line installer for macOS/Linux, PowerShell for Windows, or PyPI (`pip install octop`). Docker Compose available for production. Initialize with `octop init`, run with `octop run`, and access the web dashboard at `http://127.0.0.1:8088`.
CLI commands include `octop init`, `octop run`, `octop service start/stop`, `octop agent`, `octop channel`, `octop chats`, `octop acp`, `octop cron`, `octop models`, `octop skills`, `octop plugin`, `octop backup`, `octop clean`, `octop memory list/slim`, and `octop update`.
Configuration is managed via CLI or editing files under `~/.octop/`. Supported LLM providers include OpenAI-compatible APIs, DashScope (Qwen), Ollama, and others. IM channels include Feishu, DingTalk, QQ, Telegram, Discord, WeCom, and the web dashboard.
Security and privacy are emphasized: local-first storage, JWT multi-user isolation, PII redaction, tool approval, and user-editable shell command guardrails. No vendor lock-in for LLM providers, storage backends, or channels.
The roadmap includes shared resource pools, expert sharing, browser/terminal polishing, AgentTeams for autonomous orchestration, self-evolution, and native PC/mobile clients.
Comments
0 Rating appears after 10 ratings
Sign in to join the discussion.