Six projects and a final. One creative portfolio.
We don't watch. We ship. Every lesson produces a new small product. The capstone aggregates all 6 into one creative portfolio you submit.
What is agentic engineering?
Agentic engineering is what comes after vibe coding. Andrej Karpathy coined "vibe coding" in 2025 and retired the term a year later. The era where you prompt an agent and hope is over. The job now is to direct agents and verify what they ship.
Anyone can get an agent to produce code. Almost nobody can read it well enough to know whether it is safe to ship. That reading-and-judgment gap is the whole skill, and it is exactly where 2026 hiring sits.
Across 8 hands-on lessons you build 6 small products for real people: audit every line, write the specs the agent executes, harden the repo, then attack it. Demo Day is one creative portfolio of all 6.
"You can outsource your thinking, but you can't outsource your understanding."Andrej Karpathy, Sequoia AI Ascent 2026
Shape
Who it’s for
People who can read code already and are still in the early stretch of learning to write it. A little bit of HTML/CSS/JS or an equivalent language (Python, Java, Go, C#) is enough. Not the first programming course. Not for senior engineers. The middle, where 2026 hiring is concentrated.
You need terminal fluency, Git basics, an editor, a GitHub account, a Vercel account, and one agentic CLI installed (Claude Code, Cursor, or Codex CLI) before lesson 1.
By the end you can
- Direct an AI agent through a multi-file feature with Plan Mode and two-pass review
- Write specs the AI executes end-to-end without drifting
- Audit every line the agent ships, and apply the 3-prompt stop rule
- Harden a repo: deny-first permissions, CLAUDE.md, no leaked secrets
- Run adversarial agents against your own build, and peer-audit another team
- Ship a portfolio of six projects and a final to real people
The lessons
Capstone, graded
One creative portfolio URL plus a GitHub profile. All 6 micro-projects linked, with the audit trail visible behind each.
- 30% Portfolio creativity. A clear point of view across the 6, not just a grid of links.
- 25% Peer-audit writeup of another student's portfolio
- 20% Completeness. All 6 projects shipped, audits visible.
- 15% Adversarial validation quality on a chosen project
- 10% Portfolio polish. Load time, mobile, typography.
No other AI-coding course in 2026 makes a peer-audited creative portfolio the bar. That's the bet.
Instructor
Nika. Growth engineer, building since 14.
Last updated 2026-06-13