# learn-claude-code

> Step-by-step tutorial for building an AI agent harness from scratch

learn-claude-code is an educational guide and reference codebase demonstrating how to build an LLM agent harness from 0 to 1. Across 17 progressive lessons, it unpacks core architecture mechanisms such as the agent loop, tool dispatch, context compaction, task systems, and multi-agent coordination. Each lesson is paired with standalone Python scripts runnable via the terminal.

- Source: https://github.com/shareAI-lab/learn-claude-code
- Homepage: https://learn.shareai.run
- License: MIT
- Built with: Python 70.2%, TypeScript 29.0%, Other 0.8%
- Stars: 79589
- Forks: 12742
- Contributors: 57
- Last commit: 2026-10-27
- Purpose: AI & LLM tooling, Career builder, Agents & frameworks
- Type: CLI, Guide / course
- For: Personal

## Worth score: 71/100

How much DigGitHub recommends it, from activity, adoption, docs, license and security signals.

- Popularity: 25/25
- Momentum: 10/20
- Maintenance: 15/25
- Community: 13/15
- Readiness: 8/15

## Alternatives

- [Hermes Agent](https://diggithub.com/NousResearch/hermes-agent.md): Self-improving personal agent from Nous Research
- [nanobot](https://diggithub.com/HKUDS/nanobot.md): Ultra-light self-hosted personal AI agent
- [Ruflo](https://diggithub.com/ruvnet/ruflo.md): Orchestrate swarms of AI agents
- [OpenClaw](https://diggithub.com/openclaw/openclaw.md): Personal AI assistant that acts on your behalf across apps
- [ZeroClaw](https://diggithub.com/zeroclaw-labs/zeroclaw.md): Small, fast autonomous assistant infrastructure
- [GBrain](https://diggithub.com/garrytan/gbrain.md): Opinionated brain for OpenClaw and Hermes agents
- [qm](https://diggithub.com/yc-software/qm.md): Multiplayer agent harness for work
- [OpenManus](https://diggithub.com/FoundationAgents/OpenManus.md): Open general-purpose AI agent

---

Source page: https://diggithub.com/shareAI-lab/learn-claude-code
Updated: 2026-10-31