mini-agent

Set up a Node.js 20+ personal AI agent with Markdown memory and scheduled tasks.

3|1|Updated Feb 5, 2026
One-click install
npx skills add https://github.com/miles990/mini-agent --skill mini-agent
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mini-agent
Source: https://github.com/miles990/mini-agent/tree/main
Command: npx skills add https://github.com/miles990/mini-agent --skill mini-agent

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you set up and operate a lightweight personal AI agent that continuously uses your available context, stores long-term knowledge as Markdown, and can run proactive scheduled tasks without needing you to prompt every cycle.

Core Features & Use Cases

  • Long-term, file-based memory: Store and retrieve knowledge in Markdown so your agent’s context stays human-readable.
  • Daily notes & task tracking: Log daily conversation notes and track active/scheduled work using HEARTBEAT.md.
  • Proactivity + multi-instance support: Enable cron-driven proactive execution and run multiple isolated instances for different roles (e.g., Research vs Coding).

Quick Start

Install the agent and run it interactively by executing: curl -fsSL https://raw.githubusercontent.com/miles990/mini-agent/main/install.sh | bash, then start the assistant with mini-agent.

Frequently Asked Questions about mini-agent

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run a personal AI agent with long-term memory without using a database?

You can run a personal AI agent with long-term memory by storing knowledge as human-readable Markdown files. This file-based approach eliminates the need for databases or vector embeddings while maintaining persistent context.

How do I set up scheduled proactive task automation for a personal AI assistant?

Scheduled proactive task automation is set up using cron-driven execution, allowing the personal AI assistant to run tasks on a schedule. The agent manages active and scheduled work through a dedicated HEARTBEAT.md file.

What do I need to install a Claude CLI-based personal AI agent?

To install a Claude CLI-based personal AI agent, you need Node.js version 20 or higher and the Claude CLI available in your environment. An installer-driven workflow via a shell script automates the setup process.

Can I run multiple isolated personal AI agent instances for different roles?

Yes, you can run multiple isolated personal AI agent instances for different roles like Research and Coding. This multi-instance support allows distinct environments to operate independently with separate file-based memory.

How does file-based Markdown memory work for task tracking in AI agents?

File-based Markdown memory works by logging daily conversation notes and tracking active tasks in human-readable text files. This ensures the AI agent's context remains transparent and easily editable without database queries.