tutorial

How to Set Up Claude Code with VS Code in 2026

A step-by-step tutorial for installing Claude Code, the official Anthropic CLI, and wiring it into Visual Studio Code via the Claude Code extension. Covers npm install, authentication, extension configuration, per-project permissions, and the most common errors encountered during setup.

Overview

Claude Code is Anthropic's official CLI and IDE companion for Claude. While Claude Code runs natively in the terminal, integrating it with Visual Studio Code provides a richer editing experience: in-editor diffs, faster file navigation, and the ability to keep an existing VS Code workflow. This tutorial covers installation, authentication, IDE wiring, and common errors as of April 2026.

Prerequisites

  • Visual Studio Code 1.85 or newer
  • Node.js 20.x or newer (Claude Code is distributed as an npm package)
  • An Anthropic account with API access or a Claude Pro/Max subscription with Claude Code entitlement
  • macOS, Linux, or Windows (WSL2 recommended on Windows)

Step 1: Install Claude Code

Install the CLI globally via npm:

npm install -g @anthropic-ai/claude-code
claude --version

A successful install reports a version such as claude-code 1.x.x. If npm install fails with permission errors on macOS or Linux, fix the npm prefix to a user-owned directory rather than running with sudo:

mkdir -p ~/.npm-global
npm config set prefix ~/.npm-global
export PATH=~/.npm-global/bin:$PATH

Step 2: Authenticate

Run the auth flow:

claude auth

The CLI opens a browser window to console.anthropic.com. Sign in with the same account holding the API or Claude subscription. The browser returns a one-time code; paste it back into the terminal. Credentials are stored in ~/.claude/credentials.json with file permissions 600.

Verify auth by running:

claude --help
claude "What model am I using?"

Step 3: Install the VS Code Extension

Open VS Code and install the official "Claude Code" extension from the Marketplace:

  1. Open Extensions (Cmd/Ctrl+Shift+X)
  2. Search for "Claude Code" — publisher Anthropic
  3. Click Install

The extension auto-detects the global claude binary if it is on the PATH. If not, set the path explicitly in VS Code settings:

{
  "claudeCode.cliPath": "/Users/yourname/.npm-global/bin/claude"
}

Step 4: Open the Claude Code Panel

Open the Command Palette (Cmd/Ctrl+Shift+P) and run Claude Code: Open Chat. The panel docks in the sidebar. From here:

  • Send prompts that include the active editor file as context
  • Accept or reject diffs inline before they touch disk
  • Use / slash commands the same way you would in the terminal

Step 5: Configure Per-Project Settings

Each project can have a .claude/settings.json file that controls allowed tools, environment variables, and hooks. A minimal config:

{
  "permissions": {
    "allow": ["Read", "Edit", "Bash(git status)", "Bash(npm test)"]
  }
}

Storing this file in the repository root keeps the team aligned on which Bash commands Claude is permitted to run unattended. A CLAUDE.md file at the repo root provides project-specific instructions that the agent reads on every session.

Common Errors

  • claude: command not found — The npm global bin directory is not on PATH. Add ~/.npm-global/bin (or your npm prefix) to the shell rc file.
  • 401 Unauthorized — The auth token expired or the account lost API access. Re-run claude auth.
  • rate limit exceeded — Either the daily message budget on a Pro plan is hit, or the API tier is too low for the model selected. Check usage at console.anthropic.com.
  • VS Code extension shows "CLI not found" — Set claudeCode.cliPath explicitly. Restart VS Code after changing the setting.
  • Diffs do not apply — The extension requires the workspace to be a single root folder; multi-root workspaces can confuse path resolution. Open one folder at a time.

Tips

  • Pin the Claude Code panel to the secondary sidebar so it sits next to a terminal panel; switching context between chat and shell is faster
  • Keep CLAUDE.md in version control so onboarding new contributors does not require re-explaining repo conventions
  • For privacy-sensitive code, set ANTHROPIC_DISABLE_TELEMETRY=1 in the shell environment

Editor's Note: We use Claude Code with VS Code daily at ShadowGen for the Automation Atlas codebase and several client engagements. The VS Code extension cuts roughly 30-40% of context-switching time compared to terminal-only usage, mainly because diffs review inline with the existing tab. The biggest practical caveat is the per-project permissions.allow list — without one, the agent prompts on every shell command, which destroys flow. Spending 10 minutes upfront enumerating safe Bash commands per repo pays back within a single session. We have not yet hit a case where the VS Code extension and terminal disagree on auth state, but if it ever happens, deleting ~/.claude/credentials.json and re-running claude auth resolves it.

Written & reviewed by Rafal Fila · Last updated:

Tools Mentioned

Claude Code

Anthropic's agentic CLI tool for AI-assisted coding and automation development

AI Coding & Development Tools

Cursor

AI-powered code editor built on VS Code with native AI pair programming

AI Coding & Development Tools

GitHub Copilot

AI pair programming tool by GitHub that suggests code completions and generates functions

AI Coding & Development Tools

Related Guides

comparison

Moxo vs Lindy in 2026: Who Presses Approve When the Agent Asks

Lindy, founded in 2023, sells an AI teammate that lives in Slack and in each person's inbox; Moxo, whose predecessor Moxtra was co-founded in 2012, runs multi-party processes where approvals are steps only a person can close. Moxo wins three rounds to two: who closes the step, work across companies and the record. Lindy wins reach and testing. Vendor pages and docs read 28 September 2026.

comparison

Moxo vs Kissflow in 2026: What an Outside Approver Costs, and What the Record Proves

Moxo and Kissflow both sell approvals to people who do not write code, and each wins one round outright: Moxo's Web and Headless SDKs run the approval inside the buyer's own product, under the buyer's own sign-in. Kissflow's portal audit log stamps every event with an actor, a device and an IP, where Moxo publishes no field list. Both answers are quoted, not priced, and both meters punish a one-step request sent to hundreds of clients. Cost, the decision record, rejection routes, AI metering and governance, verified 14 to 23 September 2026.

comparison

Moxo vs Camunda 8 in 2026: BPMN User Tasks for Staff, Step Types for Outsiders

Camunda 8 can fix a process under cases already running, since active instances migrate onto a corrected definition and, for Camunda user tasks, "a migrated active user task remains assigned to the same user". Moxo keeps the shared queue that Camunda's Tasklist V2 stops evaluating, with desks that are "a shared inbox that multiple users can pick work from". Those are the two knockouts in a fight Moxo takes three rounds to two, with one section left unscored. Portability, AI review gates and a published price decide the rest, verified 14 to 23 September 2026.

Related Rankings

Best Human-in-the-Loop Automation Platforms 2026

Eight platforms are ranked here on a single mechanism: how a person enters an automated path, and what the system can prove afterward about the decision they made. Moxo scores 7.7 and ranks first; Camunda 8 scores 7.2, Microsoft Power Automate 7.1, Pega 6.9, n8n 6.7, Zapier 5.9, Kissflow 5.8 and ServiceNow 4.8, scored 22 September 2026 on the five weighted criteria. Those criteria cover the step type, the approver's account or licence, the routes available on a rejection, the audit record, and the meter that charges for the decision, and none of them can tell a review from a rubber stamp, so the methodology adds a check readers can run on any product in a few minutes: whether the decision record tells a considered approval from an instant one. Every figure was read from a vendor-owned surface and carries its own date.

Best Automation Platforms for AI Orchestration 2026

This ranking answers one question: how many real business applications can an AI agent act on out of the box? It evaluates nine platforms as of August 2026 on the reach they give an agent, not on the workflow logic they can express. That boundary is deliberate, because two neighbouring pages on this site answer different questions. Best Process Orchestration Platforms 2026 scores multi-step process control, error handling and state management. Best AI Agent Platforms 2026 scores building and hosting the agent itself. This page scores the layer between them: the connective tissue that lets an agent already built elsewhere reach the applications a business actually runs on. A platform that leads one of those pages can place low here, and two of them do. Scores derive from application and action catalogue counts, the exposure model each platform uses to publish those catalogues to an agent, setup effort, failure handling and cost per agent action. Every figure was retrieved from a vendor-owned surface on 11 August 2026 unless an earlier date is stated against it.

Common Questions

Can you automate a platform with no API using Zapier?

Not as a proper Zapier app. Zapier's help centre, updated 29 May 2026, says a private app can be built "for any service with a public API", and its fallbacks for a missing app are email parsing, RSS, webhooks, asking Zapier to add the app, or using a different app. Those let a no-API platform tell a Zap that something happened; none of them lets a Zap act inside the platform. The Zapier Agents Chrome extension can "run actions" on a page open in your own browser (help article updated 27 April 2026), but that is hands-on help, not a reusable Zap step.

How does Moxo keep humans in control when AI agents run a workflow?

Moxo keeps people on the decisions by design: approvals and other human steps are ones its product page says "only a person can close", and AI agents can fill preparer, advisor or reviewer slots around them (both read 15 September 2026). The checks on AI output are opt-in, though. In synthetic AutomationAtlas tests that day, an AI extract step's "Human review" and "Supervisor Agent" switches were both off by default, and the builder accepted the same role as a form's submitter and its approver.

How much does Moxo cost in 2026?

Moxo's only published price is Team: $5,000 a year, billed annually with no monthly option, for 100 flows and $100 of AI a year with unlimited seats (moxo.com/pricing, 28 September 2026). Scale (500 flows and $500 of AI a year) and Enterprise are custom quotes. There is no free plan, no published overage rate and no stated trial length, and the dollar AI allowance has no published conversion to the credits Moxo's product logs.

What is Moxo?

Moxo AI (app.moxo.com) is a process orchestration platform from Moxo, formerly Moxtra, for work where several parties, approvals and documents meet. You build templates of human steps, AI steps and automations, each run is a Flow with its own data and status, and outsiders act through account-free Magic Links. Its only published price is Team, and the AI agents start on the custom-quoted Scale plan (moxo.com/pricing, 15 September 2026). It is not Moxo Classic, the older app.