Build a Claude Code Verification Harness

Oct 8, Reserve a seat
Free to read

Build your own Claude Code

Frameworks hide the real engineering behind coding agents. Build your own Claude-code-style agent from scratch: tool loops, safe sandboxing, file editing, and context management, so you understand exactly what breaks in production and how to fix it.

Start readingPreview curriculum

Still deciding? Ask first.

Message a mentor about fit, prerequisites, or where to start. Replies come on WhatsApp, usually within a day.

  • Curriculum fit, prerequisites, or where to start
  • Honest answer, no pressure to enroll

Taught by an engineer who has shipped this

  • ISO 27001

    Led the engineering work behind the certification of a regulated EU platform.

  • Series A platform

    Architected the case-management product that became the business a €11.6M round was raised on.

  • 3x faster deploys

    Cut time-to-deploy by migrating to Kubernetes on GCP with deploy-on-merge.

Claude Code is the most capable AI coding agent available today. In this workshop, you reverse-engineer its architecture and build your own from scratch: the agent loop, tool design, context compaction, task management, multi-agent teams, and git worktree isolation. Each phase adds one real capability. By the end you have a production-grade agentic system.

Reverse-engineer how Claude Code works. Then build your own production AI coding agent from scratch.

What you'll ship

Real projects, not toy demos.

  • Build the core agent loop that powers Claude Code in 17 lines
  • Design safe, composable tools (bash, read, write, edit) with path sandboxing
  • Add structured planning with a TodoManager that nags the model to stay on track
  • Implement subagents with context isolation and a skill system
  • Build three-layer context compaction so your agent works indefinitely
  • Create a file-based DAG task system that survives context compression
  • Add background execution with daemon threads and a notification drain pattern
  • Build multi-agent teams with JSONL message buses and shutdown protocols
  • Implement autonomous agents that find work themselves via idle polling
  • Add git worktree isolation for safe parallel task execution

What you'll learn

You finish able to:

  • Understand the agent loop pattern that powers Claude Code and similar tools
  • Design safe, composable tools with path sandboxing and error boundaries
  • Implement three-layer context compaction for unlimited conversation length
  • Build file-based task management with dependency graphs
  • Create multi-agent teams with message buses and coordination protocols
  • Add git worktree isolation for safe parallel execution

Curriculum

From 17 Lines to a Production Coding Agent.

  1. 01

    The agent loop

    Build the core loop that powers every AI coding agent. Start with 17 lines, then add bash execution, error handling, and dual-mode operation.

  2. 02

    Tool design

    Move beyond bash-only. Design read, write, and edit tools with path sandboxing. Learn why the model IS the agent.

  3. 03

    Planning and constraints

    Plans get buried in conversation history. Add a TodoManager with constraints that keep the agent focused.

  4. 04

    Subagents and skills

    Isolate context with subagents. Externalize knowledge with skills. Optimize costs with cache-preserving injection.

  5. 05

    Memory management

    Build three-layer context compaction and file-based task graphs that survive compression.

  6. 06

    Parallelism

    Background execution with daemon threads, notification draining, and persistent teammates.

  7. 07

    Team coordination

    JSONL message buses, shutdown and approval protocols, WORK/IDLE lifecycle, and identity preservation.

  8. 08

    Worktree isolation

    The final module. Git worktrees for directory-level isolation. Tasks as control plane, worktrees as execution plane.

Who it's for

Is this for you?

AI engineers

You use agent frameworks but could not build one from scratch. That gap is becoming a liability.

Backend developers

You want to understand how tools like Claude Code, Cursor, and Copilot actually work under the hood.

Technical leads

Your team ships AI features but nobody understands the agent architecture. You need someone who can debug and extend it.

FAQ

Common questions.

  • Do I need Claude Code installed?

    No. You build your own from scratch. The workshop teaches the patterns, not the product.

  • What LLM provider do I need?

    Any provider that supports function calling. LiteLLM handles OpenAI, Anthropic, Google AI Studio (free tier), and more.

  • Should I take Building AI Agents first?

    Yes. This workshop assumes you understand tool calling and agent loops. Building AI Agents covers those foundations.

  • How long does the workshop take?

    About 14 hours. Most people finish in 3-4 sessions over a week. Each module is self-contained.