sf-shenfeng/learn-shell

Overview

An agent-native teaching OS that functions as a persistent classroom for both users and AI agents. It provides an MCP server and a self-hosted web app to facilitate interactive learning sessions. Built in TypeScript.

Best for

Best for
Developers building AI-agent teaching or training systems

Use cases

• Set up a self-hosted teaching environment for AI agents
• Create persistent learning sessions that remember context
• Integrate with MCP-compatible tools for agent-driven education

How to use

Tools exposed

• get_context
• get_learner_brief

Tested with

Claude Code, Continue

Notes

An agent-native teaching OS that functions as a persistent classroom for both users and AI agents. It provides an MCP server and a self-hosted web app to facilitate interactive learning sessions. Built in TypeScript.

3 stars on GitHub. Last updated 2026-09-07. Licensed AGPL-3.0.

Use cases

• Set up a self-hosted teaching environment for AI agents
• Create persistent learning sessions that remember context
• Integrate with MCP-compatible tools for agent-driven education

Pros

• Agent-native design integrates with AI workflows
• Self-hosted gives data control and customization
• Persistent memory across sessions

Cons

• Low adoption (3 stars) suggests limited community support
• Requires self-hosting setup and maintenance
• Niche focus may not suit traditional classroom needs

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.