AI workflows

What is Agentic coding? Definition & Context

Agentic coding is vibe coding done by an autonomous AI agent that reads context, plans steps, calls tools, and ships changes without you typing each instruction.

Agentic coding is vibe coding scaled up: instead of a back-and-forth chat, you hand a goal to an autonomous agent and it reads the codebase, plans steps, calls tools, edits files, runs tests, and reports back. The developer's role shifts to spec writing, scoping, and review — closer to running a tiny engineering team than to writing code.

The success rate depends heavily on context engineering — what files the agent can see, what tools it has, and how the spec is written. Agents are strongest on changes that are well-described, have a clear definition of done (tests pass, lint passes), and are scoped to a few files. They get worse fast on changes that span the whole architecture or rely on product taste.

Modern agentic coding setups (Claude Code, Cursor agents, Cline, OpenAI Codex CLI) all use a similar loop: read → plan → act → verify → repeat.

Read more

Related terms

Vibe coding

Vibe coding is the practice of building software by describing intent to an AI in natural language and letting the AI write, refactor, and ship the code.

Prompt engineering

Prompt engineering is the discipline of structuring instructions to language models — including context, constraints, and output format — to get reliably high-quality results.

Context engineering

Context engineering is the practice of choosing exactly what information goes into the model's context window — files, snippets, specs — so the AI has the right knowledge without noise.

Pillar page

A pillar page is the long, evergreen guide at the center of a topic cluster — broad enough to rank for the head term, with internal links out to deeper supporting articles.