CompCode Academy

Lesson 1 of 65 · Free to read

Day 1 ΓÇö What AI-assisted coding & ΓÇ£vibe codingΓÇ¥ actually are

Month 1 · Week 1 — Mental Models & Environment Setup Audience: Final-year students & early-career engineers (0–3 years) Format: Self-paced study + optional live clinic · Suggested time: 4–6 focused hours Key tools: Claude, ChatGPT, Gemini

Overview

“Vibe coding” is not magic — it is a working style: you express intent in natural language, let a model draft implementation, then you steer with constraints, tests, and taste. AI-assisted coding covers a spectrum: autocomplete (next-token help), chat (Q&A and drafts on demand), agents (goal + tool loop), and autonomous workflows (multi-step jobs with little supervision).

Today you build a personal stance as a final-year or early-career engineer. People who thrive with AI are not the ones who prompt the most — they are the ones who know when to drop autonomy and take the keyboard. That judgment shows up in capstones, internships, and interviews.

Learning objectives

By the end of this session, you can:

  • Explain the core ideas behind: What AI-assisted coding & “vibe coding” actually are; the assistance spectrum (autocomplete → chat → agent → autonomous).
  • Apply the ideas using Claude, ChatGPT, Gemini on a realistic task (capstone, internship, or first-job style)
  • Produce the exercise deliverable with clear success criteria
  • Identify common failure modes and how to recover
  • Capture at least one reusable prompt, rule, or checklist for your personal library
  • Decide what must remain human-reviewed for this class of work

Core concepts

The assistance spectrum (operational definitions)

  • Autocomplete: low agency, high locality. Best for boilerplate inside a function you already understand.
  • Chat: medium agency. Best for explanation, design options, and drafting with iterative critique.
  • Agent: high agency. The model plans, edits files, runs commands, and retries. You are the tech lead.
  • Autonomous: highest agency. Background or scheduled agents. Only appropriate with strong tests, sandboxes, and approval gates.

Vibe coding done safely

Describe the user-visible outcome, constraints (stack, style, “do not touch X”), and definition of done. Iterate visually or via diffs. Never skip reading security-sensitive code (auth, payments, data deletion).

Comparison habit

The same function spec across Claude, ChatGPT, and Gemini will differ in edge cases, naming, and error handling. Those differences teach you that models have defaults — your job is to impose yours.

Tooling for today

Primary tools: Claude, ChatGPT, Gemini. Prefer official docs when APIs change. When a model invents a symbol, verify against source or docs immediately.

Practice posture

You are a final-year or early-career engineer pairing with a very fast teammate (the model): clear intent, grounded context, inspect consequential changes, keep a paper trail (commits, prompts, decisions).

Process diagram

Watch & learn

Study these free videos before or during the exercise. Each embed is chosen to reinforce today's objectives.

Course video

Open on YouTube
Intro to Large Language Models (Andrej Karpathy) — Foundational mental model of what LLMs are — and are not.

Course video

Open on YouTube
Vibe Coding: The Complete Story (2015–2026) — Where the term came from and how practice evolved.

Worked example / case study

Scenario

Case: email validator across three models

  1. Spec: “Validate email; reject spaces; allow plus-aliases; return typed result; include 5 tests.”
  2. Send identical specs to three assistants.
  3. Score each on: correctness, edge cases, clarity, test quality.
  4. Merge the best ideas into your charter: what you will always verify manually.

Step-by-step

  1. Clarify the outcome in one sentence tied to the exercise below.
  2. Gather grounding context (files, docs, schema, acceptance checks).
  3. Run the primary tool(s): Claude, ChatGPT, Gemini.
  4. Produce the smallest vertical slice that proves the idea.
  5. Verify with a command, test, screenshot, or side-by-side comparison.
  6. Write down one steering note — what you told the AI when it went wrong.

Sample prompt you can adapt

Role: Senior engineer mentoring an early-career developer (pairing with me).
Task: What AI-assisted coding & “vibe coding” actually are; the assistance spectrum (autocomplete → chat → agent → autonomous).
Context: Final-year / early-career project (capstone, internship, or junior role).
Constraints:
- Match existing project patterns
- No new dependencies unless justified
- Include tests or a verification checklist
- Call out assumptions explicitly
Output: step plan first, then code/files, then verification commands.

Hands-on exercise

Primary task. Write a one-page personal “AI coding charter.” Give the same small function spec to three assistants and compare the outputs side by side.

Coaching notes

Charter template sections: (1) Tasks I delegate, (2) Tasks I keep, (3) Required verification, (4) Tools I prefer by task type, (5) Red lines (secrets, prod data, force-push).

Steps

  1. Set up or open the workspace required by: Claude, ChatGPT, Gemini.
  2. Skim today's core concepts and watch the video(s) before prompting.
  3. Execute the task in small commits (or saved artifacts if no-code).
  4. Fill the success checklist below.
  5. Note one surprise and one process improvement for tomorrow.

Success criteria

  • Deliverable matches the exercise description
  • You can explain every AI-generated change in plain language
  • Verification ran (tests, manual check, or published URL)
  • At least one pitfall from this lesson was consciously avoided
  • Artifact saved (repo commit, doc, or published link)

Timebox

Suggested: 90–150 minutes of focused build time after study.

Templates & checklists

AssetPurpose
Prompt stubReuse today's structure (role → task → constraints → format)
Verification checklistTests / manual QA / security skim
Decision logWhat you delegated vs kept human

Copy-paste prompt skeleton

## Goal
What AI-assisted coding & “vibe coding” actually are; the assistance spectrum (autocomplete → chat → agent → autonomous).

## Context files
- (list paths)

## Constraints
- Stack: (fill)
- Do not: (secrets, drive-by refactors)

## Definition of done
- Write a one-page personal “AI coding charter.” Give the same small function spec to three assistants and compare the outputs side by side.

Daily reflection (3 minutes)

  • What did AI do well?
  • Where did it confidently fail?
  • What rule will you add to your cookbook/rules file?

Common pitfalls

  • Declaring 'I only use agents' without a review habit
  • Treating three model outputs as interchangeable
  • Writing a charter with enthusiasm and zero boundaries

Key takeaways

  • Autonomy is a dial, not a badge of honor
  • Comparison builds taste faster than tutorials alone
  • Your charter is a living document — revisit monthly
  • Tools (Claude, ChatGPT, Gemini) change; the judgment skills transfer.

Resources

  • Official docs for: Claude, ChatGPT, Gemini
  • Course cookbook: keep prompts in a versioned Markdown file in your repo
  • Prior days this week — skim objectives if you skipped ahead

Self-paced lesson — Mastering AI-Assisted Development & Vibe Coding — Day 1 of 65.

New free lessons and course updates. Unsubscribe from any email.