What is Claude Code?
Claude Code is Anthropic's official AI coding agent. It runs in your terminal, VS Code, JetBrains IDE, or as a standalone desktop app, and it does far more than just autocomplete. Tell it to build a feature, fix a bug, write tests, or automate a tedious refactor — it plans the approach, edits the files, and verifies the result.
Unlike a simple chatbot that answers questions, Claude Code is agentic: it can read your codebase, make edits across multiple files, run shell commands, manage git commits and pull requests, and even coordinate multiple sub-agents working in parallel.
Installation
Claude Code runs on macOS, Linux, Windows, and in the browser. Pick whichever environment suits you:
- Terminal — the full CLI. Install with
curl -fsSL https://claude.ai/install.sh | bashon macOS/Linux, orirm https://claude.ai/install.ps1 | iexon PowerShell for Windows. - VS Code extension — search for "Claude Code" in the Extensions view, or install from the VS Code Marketplace. Also works in Cursor.
- Desktop app — a standalone app with visual diff review, side-by-side sessions, and scheduling. Requires a paid Claude subscription.
- Browser — no install needed. Point your browser at claude.ai/code.
- JetBrains — install the plugin from the JetBrains Marketplace.
After installing, run claude in any project directory to start a session.
Core Capabilities
Automated tedious tasks
Claude Code handles the tasks developers constantly defer: writing tests for untested modules, fixing lint errors across a project, resolving merge conflicts, updating dependencies, and drafting release notes.
claude "write tests for the auth module, run them, and fix any failures"Building features and fixing bugs
Describe what you want in plain language. Claude Code plans the approach, writes code across multiple files, and verifies it works. For bugs, paste an error message and it traces the issue through your codebase to find the root cause.
Git workflow automation
Claude Code works directly with git. Stage changes, write commit messages, create branches, and open pull requests without leaving the session:
claude "commit my changes with a descriptive message"MCP integration
The Model Context Protocol (MCP) is an open standard for connecting AI tools to external data sources. Through MCP, Claude Code can read your design docs in Google Drive, update tickets in Jira, pull data from Slack, or use your own custom tooling — all within a single coding session.
Key insight: Claude Code supports MCP servers natively. If you have a MiniMax Token Plan subscription, you can configure the MiniMax MCP server inside Claude Code to get
web_searchandunderstand_imagetools directly — no separate setup required.
Customisation: CLAUDE.md, Commands, and Hooks
Claude Code adapts to your project's conventions through a few mechanisms:
- CLAUDE.md — a markdown file in your project root. Claude Code reads it at the start of every session to understand your coding standards, architecture decisions, preferred libraries, and review checklists.
- Custom commands — package repeatable workflows as shareable commands like
/review-pror/deploy-staging. - Hooks — run shell commands before or after Claude Code actions, such as auto-formatting after every file edit or running lint before a commit.
Claude also builds auto memory as it works, saving learnings like build commands and debugging insights across sessions automatically.
Multi-Agent Teams
For complex tasks, Claude Code can spawn multiple agents that work on different parts of a problem simultaneously. A lead agent coordinates the work, assigns subtasks, and merges results — useful for large refactors or when you need several things done at once.
The Agent SDK also lets you build fully custom agents powered by Claude Code's tools and capabilities, with fine-grained control over orchestration, tool access, and permissions.
Scheduling and Automation
Claude Code is not just for interactive sessions. You can schedule recurring tasks so they run automatically:
- Cloud scheduled tasks — run on Anthropic-managed infrastructure, so they keep going even when your computer is off. Create them from the web, Desktop app, or by running
/schedulein the CLI. - Desktop scheduled tasks — run on your local machine with direct access to your files and tools.
- /loop — repeats a prompt within a CLI session for quick polling, like monitoring a build.
Perfect for morning PR reviews, overnight CI failure analysis, weekly dependency audits, or syncing docs after PRs merge.
Using Claude Code with MiniMax M2.7
You can run Claude Code with the MiniMax M2.7 model via Anthropic's compatible API. MiniMax provides an Anthropic-compatible endpoint at https://api.minimax.io/anthropic, which means Claude Code can route its requests through MiniMax instead of Anthropic's direct API — useful if you already have a MiniMax Token Plan subscription.
To configure, edit ~/.claude/settings.json and add:
{
"env": {
"ANTHROPIC_BASE_URL": "https://api.minimax.io/anthropic",
"ANTHROPIC_AUTH_TOKEN": "YOUR_MINIMAX_API_KEY",
"API_TIMEOUT_MS": "3000000",
"CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC": 1,
"ANTHROPIC_MODEL": "MiniMax-M2.7",
"ANTHROPIC_SMALL_FAST_MODEL": "MiniMax-M2.7"
}
}Make sure to clear any existing ANTHROPIC_AUTH_TOKEN or ANTHROPIC_BASE_URL environment variables before starting — these take priority over settings.json config.
Bottom line: MiniMax M2.7 inside Claude Code gives you a capable reasoning model with a 131k-token context window, running in a polished developer experience with MCP tool support, multi-agent orchestration, and built-in git workflows.
Where Claude Code Fits in the Workflow
Claude Code slots into your existing process without rewriting it. Pipe it into CI pipelines, use it for one-off bulk file operations, or keep it running as a full interactive session for complex feature work.
The same settings, MCP servers, and CLAUDE.md files work across all surfaces — terminal, VS Code, desktop, and browser. Start a task on your laptop, move it to your phone via the web interface, and hand it off to the desktop app for visual diff review when you get back to your desk.
Feature overview sourced from the official Claude Code documentation. [Source]
