Explainerwhat is vibe codingvibe coding meaningvibe coding claude code

What Is Vibe Coding? (And How to Do It Without Babysitting)

Vibe coding is letting an AI agent drive while you steer. Here's what the term actually means, where it works, where it breaks, and how to do it without staring at a terminal all day.

Ian Brillantes · Founder & iOS EngineerJune 13, 20265 min read

Quick answer

Vibe coding is building software by describing what you want in plain language and letting an AI agent write the code while you steer through feedback instead of typing it yourself. You shift from author to director: you own intent, direction, and review, and the agent handles the mechanics of editing files and running commands.

"Vibe coding" went from a half-joke to a real workflow in about a year. The term gets thrown around loosely, so here's a straight answer: what it actually means, where it genuinely works, where it falls apart, and how to do it without chaining yourself to a terminal.

The actual definition

Vibe coding is building software by describing what you want in plain language and letting an AI agent write the code, while you steer through feedback rather than typing most of it yourself. You stay in the loop on intent and direction; the agent handles the mechanics.

The phrase was popularized in early 2025 by Andrej Karpathy, describing a mode where you "fully give in to the vibes" and mostly accept what the model produces, nudging it when something's off. The key shift: you move from author to director.

Info

Vibe coding is a spectrum, not a switch. On one end you read every line; on the other you barely glance at the code and judge by whether the app works. Most productive sessions sit in the middle: loose on mechanical work, tight on anything that matters.

What it looks like in practice

A typical session with a tool like Claude Code:

  1. You describe a feature or a bug in a sentence or two.
  2. The agent reads the relevant files, proposes a plan, and starts editing.
  3. It runs commands — tests, installs, builds — pausing to ask permission for anything consequential.
  4. You review the diff, react ("good, but handle the empty case"), and it iterates.

You're reading and reacting far more than you're typing code. That's the vibe.

Where it genuinely shines

  • Greenfield and prototypes — going from nothing to a working thing fast, where there's no legacy to respect.
  • Well-scoped, mechanical work — migrations, test coverage, boilerplate, renaming across a codebase.
  • Unfamiliar territory — a language or framework you don't know well, where the agent knows the idioms.
  • Parallel work — kicking off several agents on independent tasks at once (more on running agents in parallel).

Where it breaks

Vibe coding is a tool, not a religion. It struggles when:

  • The stakes are high. Security-sensitive code, payments, anything where a subtle wrong assumption is expensive. Review gets non-negotiable.
  • The codebase is large and subtle. Deep, interlocking systems where the right change depends on context the agent can't fully see.
  • You stop reading the diffs. The failure mode isn't the AI writing bad code — it's you merging code you never understood.
Heads up

The honest rule: vibe code the first draft, but own the review. The agent is accountable for nothing. You stay accountable for what ships, so the diff is the one thing you never skip.

The hidden cost no one mentions

Here's the part that surprises people: the bottleneck in vibe coding isn't the AI's speed. It's your availability. The agent constantly stops — to ask permission, to confirm a direction, to report it's done. Every one of those pauses is dead time if you're not watching.

So you end up babysitting: staring at a terminal, waiting for the next prompt, afraid to step away because the run will just sit there. That defeats the entire point. The whole promise was that the agent does the work so you don't have to hover.

How to vibe code without babysitting

The fix is to make the agent come to you instead of you watching it. Claude Code hooks fire on every meaningful event, so you can route them to your phone and only engage when there's an actual decision to make:

  • Start the agent, then close the laptop lid in your head — walk away.
  • Get a push notification the instant it needs a permission or finishes a task.
  • Glance at a Live Activity on your lock screen to see who's cooking and who's waiting.

That turns vibe coding from "trapped at the desk" into what it was supposed to be: you set direction, the agent does the work, and you step in only at the moments that matter.

Stop babysitting your terminal

Agentfy pushes Claude Code status to your iPhone — Live Activities, Dynamic Island, and instant alerts.

Download Agentfy

The setup is hooks-only and takes a couple of minutes on iOS 18+: install the open-source Agentfy plugin (or paste its setup prompt) and it wires up the hooks for you. Once it's wired up, you can leave a run going while you grab coffee, ride the train, or start a second agent — and still catch the exact moment one of them needs you. That's the difference between vibe coding in theory and vibe coding you can actually live with. If you want the full walkthrough, see Claude Code on iPhone.

About the author

Ian Brillantes · Founder & iOS Engineer

Ian is the founder of Agentfy and a senior software engineer who lives in Claude Code daily. He builds the hooks-to-push-notification pipeline behind Agentfy and writes these guides from the same multi-agent workflow they describe.

Part of Running Claude Code Like a Pro

More guides in this series

Frequently asked questions

Is vibe coding good or bad?+
Vibe coding is neither inherently good nor bad; it is a tool that fits some jobs and fails others. It shines for prototypes, boilerplate, and unfamiliar stacks where speed matters. It gets risky on high-stakes code or large, subtle systems, especially if you stop reading the diffs and merge code you never actually understood.
Who coined the term vibe coding?+
The phrase was popularized by Andrej Karpathy in early 2025. He described a mode where you fully give in to the vibes, mostly accept what the model produces, and nudge it when something is off. The term spread quickly because it named a workflow many developers were already drifting into with AI coding agents.
What is the difference between vibe coding and traditional coding?+
In traditional coding you type most of the code yourself and the editor is a passive tool. In vibe coding you describe intent in plain language and an AI agent writes and runs the code while you review and redirect. You move from being the author of every line to being the director who owns intent, scope, and final review.
Is vibe coding safe for production code?+
It can be, but only if you keep ownership of review. The safe rule is to vibe code the first draft and then read every diff before it ships, paying extra attention to security, payments, and edge cases. The failure mode is not the AI writing bad code; it is a human merging code they never understood.
Do I need to watch the terminal the whole time when vibe coding?+
No, and you shouldn't have to. The real bottleneck is your availability, because the agent keeps pausing for permissions and confirmations. Routing Claude Code hooks to push notifications and Live Activities lets the agent come to you, so you only step in at the moments that need a decision instead of hovering over a terminal.

Ready to stop babysitting your terminal?

Agentfy pushes Claude Code status to your iPhone — Live Activities, Dynamic Island, and instant alerts. Set up in under a minute.

Download Agentfy