Claude Code 1.0: Everything You Need to Know About the CLI Revolution
Claude Code just hit 1.0 with a redesigned CLI, persistent memory, and MCP tool support. Here's what matters.
In this article
What's Actually New in 1.0
The headline features are persistent memory (Claude remembers project context across sessions), first-class MCP tool support (connect any tool server and Claude can use it), and a redesigned permission system that lets you pre-approve categories of actions rather than clicking 'allow' on every shell command. Together, these three changes move Claude Code from a clever demo to a tool you can leave running on a real task.
Early benchmarks from Anthropic's internal teams show a 3.2x improvement in task completion speed compared to the previous beta, largely driven by the memory system eliminating re-explanation overhead.
The MCP Integration Story
The Model Context Protocol (MCP) is an open standard for connecting AI tools to external data sources. Claude Code 1.0 makes MCP a first-class integration layer. You can connect Claude to your design docs in Google Drive, update tickets in Jira, pull context from Slack, or wire in any custom tooling your team already uses. Each MCP server is defined in a local config file, and Claude Code picks it up automatically at session start.
{
"mcpServers": {
"github": {
"command": "npx",
"args": ["-y", "@anthropic-ai/mcp-github"]
}
}
}Important: MCP servers run locally on your machine. Claude Code never sends your tool configurations or credentials to Anthropic's servers.
We built Claude Code because we wanted an AI assistant that could actually do the work, not just suggest edits you have to apply manually. (Anthropic Engineering Blog)
Where Claude Code Runs
Claude Code 1.0 is not a terminal-only tool. The same session and configuration layer works across five environments. In the terminal, you install via a one-line curl command on macOS, Linux, or WSL, or via PowerShell on Windows. WinGet and Linux package managers (apt for Debian, dnf for Fedora, apk for Alpine) are also supported. The VS Code and Cursor extensions add inline diffs, at-mentions, and conversation history directly in your editor. The JetBrains plugin covers IntelliJ IDEA, PyCharm, and WebStorm with interactive diff viewing. A standalone desktop app for macOS (Intel and Apple Silicon) and Windows lets you review diffs visually and run multiple sessions side by side. Finally, the web interface at claude.ai/code requires no local setup, making it practical for reviewing a repo you have not cloned locally.
CLAUDE.md and Auto-Memory
CLAUDE.md is a markdown file you add to your project root. Claude Code reads it at the start of every session and treats its contents as standing instructions. Use it to define coding standards, record architecture decisions, list preferred libraries, and maintain review checklists. Teams can commit CLAUDE.md to version control so every contributor gets the same baseline context without a verbal handoff. Beyond the explicit file, Claude Code 1.0 also builds auto-memory as it works: it saves learnings like build commands, test runner flags, and debugging patterns across sessions without you writing anything. Both systems together mean you explain your project once and Claude carries that knowledge forward.
Skills, Hooks, and Customization
Skills are packaged, repeatable workflows your team can share. You define a skill once and it becomes a slash command anyone can call: /review-pr runs your standard pull request checklist, /deploy-staging triggers your staging pipeline. Hooks are shell commands that fire automatically in response to Claude Code events. A common pattern is auto-formatting after every file edit and running lint before a commit, catching style issues without a manual step. Together, skills and hooks let you encode your team's workflow norms into the tool rather than relying on everyone remembering the conventions.
Team Workflows and CI/CD
Claude Code 1.0 integrates directly with GitHub Actions and GitLab CI/CD for automated PR review and issue triage. You can route bug reports from Slack and receive a pull request back without leaving the chat. For parallel work, you can spawn multiple Claude Code agents that each handle a different part of the task simultaneously, with a lead agent coordinating and merging results. Routines run scheduled tasks on Anthropic-managed infrastructure, so they keep executing even when your computer is off: triggered on a schedule, by API call, or by GitHub events. For custom automation beyond what the built-in surfaces cover, the Agent SDK gives you full control over orchestration, tool access, and permissions.
Safe Commands and Permission Control
A key friction point in previous developer previews was command anxiety. Version 1.0 resolves this by introducing fine-grained control policies. Developers can toggle interactive shells, set permanent read-only scopes for sensitive subdirectories, and configure automatic overrides. Additionally, all shell commands are logged with full audit trails, giving teams the visibility they need before trusting an agent with production systems.
Configuration and Workspace Integration
Setting up Claude Code 1.0 takes less than a minute. The utility automatically reads local config profiles and environment flags. By modifying the global settings.json file, developers can bind custom aliases, configure default project templates, and pre-load specialized prompts that enforce team-specific conventions. Sessions are not tied to a single surface: you can start work on the web, continue on mobile via Remote Control, and pull it into your terminal with the --teleport flag.
Should You Upgrade?
If you are using the Claude Code beta today, upgrading is a clear call. The memory system alone eliminates the biggest friction point: having to re-explain your project every session. Combined with MCP tools, you can build genuinely autonomous workflows that handle research, coding, testing, and deployment in a single session. Native installs update automatically in the background, so you stay current without a manual upgrade step. The only thing to weigh is whether your team is ready to commit CLAUDE.md to your repos: once you do, the productivity compounding is significant.
Sources and Further Reading
Full documentation and feature reference are available at docs.anthropic.com/en/docs/claude-code
Get the Claude playbook in your inbox.
One weekly email for Claude and Claude Code users. Real workflows, no hype. Subscribe and we send you The Claude Power-User Cheatsheet.
— ¶ —
Luke Thompson
Luke Thompson is the founder of The Operations Guide, LLC and editor of The Claude Insider. Based in Jonesborough, Tennessee, he has spent years building AI-augmented business systems and automation workflows for operators and teams. He began working with large language models in production well before the current wave of consumer AI tools, integrating them into client workflows, content pipelines, and operational infrastructure. At The Claude Insider, he writes about Claude with the specificity of someone who uses it daily as a professional tool — not as a reviewer or commentator, but as a builder. His coverage focuses on what actually works: prompt patterns, API integration strategies, agentic workflows, and the real-world tradeoffs that practitioners face. He is not affiliated with Anthropic, PBC.
Articles are researched and drafted with AI assistance, reviewed and edited by Luke Thompson.
Know where AI can pay off in your company.
Take the free two-minute AI Readiness Assessment. See your score, the two gaps holding you back, and the next move worth making.
