Last verified

Is Claude Code an agent?

Yes.
Acts and iterates · not a chatbot

Yes — Anthropic calls it an "agentic coding tool", and it passes the operational test rather than just the marketing one: given a goal it decides which files to open, edits several, runs your commands, reads the output and tries again, without a prompt per step. It also spawns subagents that split a task in parallel. What it isn't is autonomous — it asks permission before acting.

Plans then acts — multi-file edits from a single goal Runs and reacts — executes tests, reads failures, iterates Subagents — several sessions split a task, a lead merges Not autonomous — permission model gates real actions Not a chatbot — it changes files, not just answers
§ 01 / SOURCE

Where this answer comes from.

Methodology →
PER VENDOR DOCS + OUR RUNS · AUG 01, 2026

The label is the vendor's; the behaviour is what settles it.

Anthropic's own definition opens with "an agentic coding tool that reads your codebase, edits files, runs commands, and integrates with your development tools". Labels are cheap, so the more interesting evidence is the shape of what it does: the docs' own example prompt is "write tests for the auth module, run them, and fix any failures" — one instruction covering discovery, authorship, execution and repair.

Two features only an agent needs. Subagents: several Claude Code agents work on different parts of a task while a lead coordinates and merges. The Agent SDK: build your own agents on the same tools and orchestration. Neither makes sense for a completion tool.

And a limit that matters more than the label. It is not autonomous: by default it asks before changing files, and it asked us for approval before running a script and before searching the web. Agency here means deciding what to do, with a human still gating the doing.

Vendor's term: agentic coding toolDecides steps: yesRuns commands: yes — reads outputParallel subagents: yesAutonomous: no — permission-gated
!
Per vendor docs — we did not run this test. The answer above rests on the source below, read on . Where we run a capability ourselves, this section carries the terminal transcript instead.
Source Anthropic docs — Claude Code overview https://code.claude.com/docs/en/overview
§ 02 / WHY

What makes it agentic.

2.1
The test is who decides the next step

A completion tool answers the question in front of it. An agent chooses the next question. That's the whole distinction, and it's visible the first time you give Claude Code a goal instead of an instruction: it goes looking, forms a plan, and comes back having changed several files.

The feedback loop is the second half. Running your tests and reading the failure is what separates an agent from a very good code generator — it can tell whether it succeeded.

2.2
Agency stops at the permission boundary

Worth being precise, because "agent" makes people imagine something unsupervised. In default mode it proposes and asks. We watched it decline to search the web until the tool was approved, and offer to write a script rather than silently running one. There are modes that loosen this, up to a flag whose name is a warning label.

So: autonomous in its reasoning, supervised in its effects. If you want it unsupervised in CI, that's a deliberate configuration rather than the default.

2.3
Agents plural

Beyond a single session, it can spawn subagents that take different parts of a task simultaneously, with a lead agent assigning work and merging results, and background sessions you can watch from one screen. The Agent SDK exposes the same machinery for building your own.

That's the part where the word stops being a marketing adjective and starts describing an architecture.

→ SAME CATEGORY

Other agentic coding tools.

The category now has several credible entries; they differ on where the agent runs and what it costs rather than on whether they qualify.

Frequently asked.

Quick follow-ups people search after this question.

Q · 01 What's the difference between an AI agent and a chatbot? +
A chatbot answers; an agent acts. Claude Code decides which files to open, edits them, runs commands and reads the results to decide what's next. The output is a changed repository, not a message.
Q · 02 Does it work autonomously? +
Not by default. It asks before changing files and before running things — we saw it decline to search the web until the tool was approved. Looser permission modes exist, including one whose name is explicitly a warning.
Q · 03 Can it run several agents at once? +
Yes — subagents split a task in parallel with a lead session merging results, and background sessions can be watched from one screen. The Agent SDK exposes the same orchestration for custom agents.
Q · 04 Is it an AI agent or an assistant? +
Both words get used loosely, but the meaningful line is who picks the next step. Here it does. That makes it an agent, with a permission boundary keeping a human in the loop for effects.