Skip to content

MiQi

A lightweight, extensible personal AI agent framework for production automation and conversational workflows.

Python Status License Ruff Docker

MiQi is a compact AI agent runtime designed for developers who want a self-hosted, programmable assistant. It connects to any OpenAI-compatible LLM provider and exposes a rich toolset — file operations, shell execution, web search, scheduled tasks, sub-agents, and external MCP servers — all configurable via a single JSON file.

Features

Category Capabilities
LLM Providers OpenRouter, OpenAI, Anthropic, DeepSeek, Gemini, Groq, Moonshot, MiniMax, ZhipuAI, DashScope (Qwen), SiliconFlow, VolcEngine, AiHubMix, vLLM, Ollama, OpenAI Codex (OAuth), and any OpenAI-compatible endpoint
Built-in Tools File system, shell, web fetch/search, paper research (search/details/download), cron scheduler, sub-agent spawning
Channels Feishu is wired in the packaged gateway today; additional adapter modules for Telegram/Discord/Slack/Email/QQ/DingTalk/MoChat are present in the repository
MCP Integration Connect any MCP-compatible tool server; seven domain-specific MCP servers bundled
Memory RAM-first snapshots, self-improvement lessons, and workspace-local JSONL session history
Agent Runtime Concurrent tool execution for safe batches, iteration-budget safeguards, queue notifications, and MCP heartbeat progress
Extensibility MCP server integration, skill files, custom provider plugins
CLI Interactive onboarding, agent chat, gateway mode, cron and memory management

Quick Navigation

  • :material-rocket-launch: Getting Started — Installation, quick start, and first run
  • :material-cog: Configuration — Config file reference and environment variables
  • :material-console: CLI Reference — All commands, tools, and their options
  • :material-puzzle: MCP Integration — Connect bundled and custom MCP servers
  • :material-sitemap: Architecture — System design and module breakdown
  • :material-brain: Memory System — RAM-first memory architecture
  • :material-school: Self-Improvement — Lesson extraction and feedback loop
  • :material-shield: Security — Security policy and best practices
  • :material-code-braces: Developer Guide — Contributing and development setup

Acknowledgements

MiQi is a domain-focused evolution of the upstream nanobot project. Full credit to the upstream team for the excellent engineering baseline in runtime design and tool abstraction.

Reference baseline: nanobot @ 30361c9 (2026-02-23)