---
slug: "claude-code-mastery-chaiwithjai"
source_type: "readme"
source_url: "https://cdn.jsdelivr.net/gh/chaiwithjai/claude-code-mastery@main/README.md"
repo: "https://github.com/chaiwithjai/claude-code-mastery"
source_file: "README.md"
branch: "main"
---
# Claude Code Mastery
**Go from "AI autocomplete user" to "AI agent architect" in one livestream.**
---
## The Problem
You use Cursor. Or Copilot. Or both. You've gotten faster at writing code.
But you're still just using autocomplete with extra steps.
- Your `.cursorrules` file is ignored half the time
- Context resets every session
- You're prompting, not architecting
- AI helps you code, but doesn't know YOUR stack
**You're using AI. You're not building with AI.**
This course fixes that.
---
## The Transformation
```
CURSOR/COPILOT CLAUDE CODE
───────────────────────────────────────────────────────────────────
Single-file autocomplete → Multi-file agents with memory
"Chat in sidebar" → Skills that encode YOUR expertise
.cursorrules (ignored) → CLAUDE.md (sacred architecture)
Generic AI assistant → Domain expert that knows your stack
Every session starts fresh → Compound knowledge across sessions
───────────────────────────────────────────────────────────────────
```
**Duration**: 3.5 hour livestream (built in public)
**Outcome**: Your first Claude Code skill shipped, production setup complete
---
## What You'll Build
| Segment | You'll Ship | Time |
|---------|-------------|------|
| [0. Cold Open](https://github.com/chaiwithjai/claude-code-mastery/blob/HEAD/segments/00-cold-open.md) | See a working agent demo | 5 min |
| [1. Architecture](https://github.com/chaiwithjai/claude-code-mastery/blob/HEAD/segments/01-architecture.md) | CLAUDE.md for your project | 20 min |
| [2. First Skill](https://github.com/chaiwithjai/claude-code-mastery/blob/HEAD/segments/02-first-skill.md) | Complete skill with SKILL.md | 40 min |
| [3. Plugin Architecture](https://github.com/chaiwithjai/claude-code-mastery/blob/HEAD/segments/03-plugin-architecture.md) | Agent + command | 35 min |
| [4. MCP Integration](https://github.com/chaiwithjai/claude-code-mastery/blob/HEAD/segments/04-mcp-integration.md) | External API connection | 25 min |
| **Break** | Catch up | 10 min |
| [5. Telemetry](https://github.com/chaiwithjai/claude-code-mastery/blob/HEAD/segments/05-langsmith-telemetry.md) | LangSmith tracing | 20 min |
| [6. Guardrails](https://github.com/chaiwithjai/claude-code-mastery/blob/HEAD/segments/06-git-guardrails.md) | Safety constraints | 15 min |
| [7. Domain Expertise](https://github.com/chaiwithjai/claude-code-mastery/blob/HEAD/segments/07-domain-expertise.md) | Your framework as a skill | 25 min |
| [8. Subagents](https://github.com/chaiwithjai/claude-code-mastery/blob/HEAD/segments/08-subagent-patterns.md) | Specialist delegation | 15 min |
| [9. Production Setup](https://github.com/chaiwithjai/claude-code-mastery/blob/HEAD/segments/09-production-checklist.md) | Complete, verified setup | 10 min |
| [10. Team Adoption](https://github.com/chaiwithjai/claude-code-mastery/blob/HEAD/segments/10-team-adoption.md) | Rollout playbook | 30 min |
Every segment ends with something you built yourself.
---
## Who This Is For
**This course is for you if:**
- [ ] You use Cursor, Copilot, or similar AI coding tools
- [ ] You want to go beyond autocomplete to building agents
- [ ] You have domain expertise worth encoding (frameworks, patterns, processes)
- [ ] You want your team to adopt AI-assisted development systematically
**This course is NOT for you if:**
- You've never used AI coding tools (start with basics first)
- You just want faster autocomplete (Cursor is fine for that)
- You don't have a specific domain or codebase to apply this to
---
## Why This Works
### 1. Live Build, Not Lecture
I build everything live during the stream. You follow along. No slides-only segments.
### 2. Real Patterns from Production
Every pattern comes from my actual Claude Code setup: 17 custom skills, multiple plugins, LangSmith traces showing real workflows.
### 3. Encode YOUR Expertise
By the end, you've encoded your own framework/process as a Claude Code skill. Not a toy example—your actual workflow.
### 4. Team-Ready
The final segment is a complete team adoption playbook. This isn't just personal productivity—it's how you roll out to your org.
---
## The Evidence
This course uses exported LangSmith traces to show exactly how Claude Code works:
| Trace | What It Shows |
|-------|---------------|
| [Skill Activation](https://github.com/chaiwithjai/claude-code-mastery/tree/HEAD/traces/) | How skills load and route |
| [Agent Workflow](https://github.com/chaiwithjai/claude-code-mastery/tree/HEAD/traces/) | Multi-step agent execution |
| [MCP Integration](https://github.com/chaiwithjai/claude-code-mastery/tree/HEAD/traces/) | External API calls |
| [Subagent Delegation](https://github.com/chaiwithjai/claude-code-mastery/tree/HEAD/traces/) | Specialist handoff |
Not conceptual diagrams. Actual traces from production use.
---
## Quick Start
### Before the Livestream
```bash
# 1. Install Claude Code
# Follow: https://claude.ai/code
# 2. Clone this repo
git clone https://github.com/ChaiWithJai/claude-code-mastery.git
cd claude-code-mastery
# 3. Complete pre-setup
cat SETUP.md
```
### During the Livestream
Follow along with each segment. At pause points, build your own version.
### After the Livestream
Complete any exercises you missed. Use the reference guides for lookup.
---
## What Success Looks Like
### By End of Livestream
- [ ] CLAUDE.md documenting your project architecture
- [ ] One working skill with SKILL.md
- [ ] One command added to your skill
- [ ] One domain expertise skill draft
- [ ] Team adoption plan outline
### Within 30 Days
- [ ] Full production setup verified
- [ ] LangSmith tracing active
- [ ] At least 3 custom skills in daily use
- [ ] Team rollout initiated
### Within 90 Days
- [ ] Your team uses Claude Code systematically
- [ ] Custom skills for your domain
- [ ] Measurable productivity gains
**[Self-Assessment Rubric →](https://github.com/chaiwithjai/claude-code-mastery/blob/HEAD/assessments/mastery-rubric.md)**
---
## For Teams
**Want to upskill your entire engineering team?**
This course works for:
- **Platform teams** building internal developer tools
- **Engineering orgs** standardizing on AI-assisted development
- **Startups** wanting to maximize developer leverage
- **Agencies** building AI-powered workflows for clients
**[Team Adoption Guide →](https://github.com/chaiwithjai/claude-code-mastery/blob/HEAD/FOR-TEAMS.md)**
---
## Repository Structure
```
claude-code-mastery/
├── segments/ # Livestream content (11 segments)
├── exercises/ # Pause point exercises (5 folders)
├── examples/ # Reference implementations
│ ├── skills/ # 3 example skills
│ ├── agents/ # Agent patterns
│ ├── commands/ # Command patterns
│ └── plugins/ # Minimal plugin example
├── traces/ # LangSmith exports
├── references/ # Quick reference guides
├── assessments/ # Self-evaluation rubric
├── FOR-TEAMS.md # Team adoption playbook
└── COMPLETION.md # Certificate + LinkedIn templates
```
---
## The Philosophy
### "The Soul Lives in Prompts"
Claude Code isn't about better autocomplete. It's about encoding your expertise.
Your system prompts should be 100-500 lines of domain knowledge. Your skills should make Claude an expert in YOUR stack. Your agents should know YOUR workflows.
Generic AI assistants are commodities. Domain experts are valuable.
### CLAUDE.md is Sacred
`.cursorrules` is a suggestion. `CLAUDE.md` is architecture documentation.
When you write a CLAUDE.md, you're not configuring an AI tool. You're documenting your system for any intelligent reader—human or AI.
### Skills Compound
Every session with Cursor starts fresh. Every skill in Claude Code compounds.
Build once, use forever. Your expertise becomes infrastructure.
---
## FAQ
**How is this different from Cursor/Copilot?**
Cursor/Copilot are autocomplete tools. Claude Code is an agent platform. You can build skills, agents, commands, and integrations that persist across sessions and encode your domain expertise.
**Do I need to uninstall Cursor?**
No. Many developers use both. Claude Code for architecture and agents, Cursor for in-file editing. They complement each other.
**What if I can't attend live?**
The repo contains everything. Watch the recording, work through segments at your own pace, use exercises for practice.
**Is this for beginners?**
No. This assumes you already use AI coding tools and want to go deeper. If you've never used Cursor/Copilot, start there first.
---
## Ready?
The gap between "AI user" and "AI architect" is smaller than you think.
**[Start with Setup →](https://github.com/chaiwithjai/claude-code-mastery/blob/HEAD/SETUP.md)**
---
## Work With Me
**For individuals**: This free course covers the full livestream content. If you want personalized coaching on encoding your specific domain expertise, **[reach out](https://linkedin.com/in/chaiwithjai)**.
**For teams**: Team packages include custom skill development for your codebase, live workshops, and adoption tracking. **[Contact for team pricing](mailto:jai@chaiwithjai.com?subject=Claude%20Code%20Mastery%20-%20Team)**.
---
Built by Jai Bhagat
Agent Engineer (LLM Observability + Evaluations) • Instructional Designer • On a mission to teach 20 million people
Star this repo if you ship your first skill.