INITIALIZING BUNKROS IDENTITY LAB
LOC UNDERGROUND
SYS --:--:--
Learning / Code Generation

Speak logic, get software.

You don't need to memorise syntax to build things any more. With AI you describe the logic in plain English and the machine writes the code — and helps you debug it. But the real skill isn't typing prompts; it's validating, securing and orchestrating what comes back. This class takes you from typist to architect.

You'll be able to

Describe, not type

Turn plain-language logic into working components.

You'll be able to

Debug with AI

Use the model as a tireless pair-programmer and reviewer.

Watch for

Confident wrong code

AI writes plausible bugs and imaginary packages. Verify everything.

01 — Warm-up

True or false?

"If AI-written code compiles and the demo works, it's safe to ship." True or false?

Self-Check · True / False
02 — Tools

The dev toolkit

You don't need an engineering degree to start — you need to know the shapes of the tools. Most people run a hybrid: a fast in-editor assistant for everyday edits plus an autonomous agent for bigger jobs.

AI editor

Cursor

An AI-native code editor — inline suggestions, codebase-aware chat and an agent mode. A common daily driver.

Agent

Claude Code

A terminal agent for multi-step jobs; extensible via MCP to reach your tools and data.

Everywhere

GitHub Copilot

Solid completion and chat across many editors; best value entry point, wired into GitHub.

Pair

A chat model

Claude/GPT/Gemini as a pair-programmer for reasoning, explaining and planning before you write.

03 — Context

Context engineering & MCP

An AI is only as good as what it can see. The biggest quality lever isn't a clever prompt — it's getting the right code, docs and data into the model's window.

Token discipline

Feed what's relevant

Don't dump the whole repo. Give the relevant files and interfaces; summarise the rest.

MCP

Model Context Protocol

An open standard connecting the model to external tools — APIs, databases, live docs — turning a chatbot into an agent wired into real systems.

Project memory

Rules files

Give the tool a standing brief — conventions, do/don't — so every session starts aligned.

Big repos

Retrieval & indexing

Use the tool's index so it pulls the right symbols on demand instead of guessing.

04 — Interactive

English → code

Click a plain-English request and watch the AI "type" working code. Notice how a clear, specific request maps cleanly to structured code — vague intent would produce vague output.

Sandbox · Describe the logic, read the code
# Pick a request above — the AI will write the code.
05 — Method

Agentic coding & TDD

Autonomy

Agentic workflows

Agents can plan, edit many files, run commands and check their own work. The clearer your spec and acceptance criteria, the better they perform — vague intent produces confident nonsense.

TDD

Tests first

Ask for the tests before the code — edge cases and integration paths. The tests become the spec the AI must satisfy, and your proof it worked.

Vibe coding

Natural-language logic

Describe behaviour in plain text to scaffold components fast — then review and harden like any code.

Debugging

AI as detective

Paste the stack trace and repro steps; let the model form hypotheses and propose minimal fixes you confirm.

Core insight

Agentic coding shifts your effort from writing to specifying and supervising. Your value is knowing whether the output is right, safe and well-built — and owning that call.

06 — In the Wild

Refactoring a legacy script

Priya volunteers for a mutual-aid group and inherited a tangled donation-tracking script no one understands.

Before

"Just fix it"

She asked an agent to "make this better." It rewrote everything at once, broke the CSV export, and she couldn't tell what changed.

After

Spec + tests + small steps

She had the AI write tests for the current behaviour first, then refactor in small steps behind those tests, reviewing each diff.

Result: cleaner code, the export still works, and she can explain every change.
Why it matters

The win wasn't the AI's cleverness — it was Priya's process: tests as a safety net, incremental change, and a human reviewing every diff. That's engineering; the AI just typed faster.

07 — Research

Pick your setup

Tools shift fast. Find which AI coding setup fits a beginner and note one safety practice it encourages.

Search for

What to look up

"Cursor vs Copilot vs Claude Code" · "AI coding for beginners" · "MCP servers list".

Judge the source: recent comparisons and the makers' docs; check the date — this space changes monthly.
Your findings
Saved locally on your device.
08 — Assignment

Write an agentic brief

Turn a vague request — "build me a sign-up form" — into a strong brief: the stack, the requirements, tests first, and acceptance criteria, with a dependency guardrail.

Your brief
Saved locally.
Success rubric — tap to expand

Needs work: repeats the vague request.
Getting there: adds a stack but no tests or acceptance criteria.
Solid: stack + requirements + tests, but no dependency guardrail.
Excellent: stack, security requirements, a TDD order, a "flag new dependencies" rule, and clear acceptance criteria.

09 — Retrieval practice

Quiz — 5 questions

Question 1 · Medium
What most improves an AI coding agent's output?
A
Saying "please" and "thank you".
B
Pasting the entire repo every time.
C
Engineering the context — feeding the right files, interfaces and docs.
Question 2 · Medium
In test-driven development with AI, what comes first?
A
The deployment pipeline.
B
The tests — they define what "working" means.
C
The marketing copy.
Question 3 · Medium
An agent tells you to install a package you've never heard of. First move?
A
Verify it actually exists and is the real, maintained package before installing.
B
Install it immediately — the agent wouldn't be wrong.
C
Disable the build error so it stops complaining.
Question 4 · Easy
What is MCP?
A
A programming language.
B
An open standard connecting a model to external tools, data and docs.
C
A type of GPU.
Question 5 · Easy · True / False
"If it compiles and the demo works, it's ready to merge." True or False?
T
True
F
False
10 — Security & mistakes

Spot the fake package

AI sometimes invents plausible-but-fake dependencies — and attackers register those names with malware ("slopsquatting"). One of these install lines points to a package that doesn't really exist. Click the one you'd never install without checking.

Mini-game · Which package is hallucinated?
pip install requests
npm install express
pip install fastjson-turbo-secure
npm install lodash
Trap

Trusting confident code

Fluent, well-formatted code can still be wrong. Fix: run it and read it — confidence is not correctness.

Trap

Auto-installing deps

Running suggested installs blind. Fix: verify every package is real, maintained and widely used.

Trap

Big-bang rewrites

Letting an agent rewrite everything at once. Fix: refactor incrementally behind tests.

Trap

Skipping review

Merging unread AI diffs. Fix: a human reviews and owns the merge — always.

Debug your thinking — before you merge
11 — Reflect

Typist to architect

If anyone can now generate code, what becomes more valuable — and what becomes less? Where would you refuse to ship AI code without a human who deeply understands it?

Private journal
Saved locally to your device.
Well done

You can direct code now — and, more importantly, judge it. Next, we tackle the hardest medium: generative video.

12 — Reference

Glossary — precise words

AI coding agent
A tool that plans and executes multi-step coding tasks with minimal step-by-step input.
Context engineering
Curating the right code, docs and data into the model's window for accurate output.
MCP
Model Context Protocol — an open standard linking models to external tools and data.
Vibe coding
Generating software from natural-language descriptions rather than hand-written syntax.
TDD
Test-driven development — writing tests before code; a natural fit for AI.
Refactoring
Improving code structure without changing behaviour — incrementally, behind tests.
Hallucinated package
A plausible but non-existent dependency an AI suggests; a supply-chain risk.
Stack trace
The error report showing where code failed — gold for AI debugging.
Pull request
A proposed code change for review before it merges.
Scaffolding
Generating the skeleton of an app or feature to build on.
CI/CD
Automated build, test and deploy pipelines.
Architect mindset
Shifting from writing code to specifying, securing, reviewing and orchestrating it.