Writing for operators and engineers who want to go deep on Claude.
Tutorials, comparisons, and opinionated takes on Claude Code, the Agent SDK, the Messages API, MCP, and how real teams ship with Claude in production. No fluff, no hype.
We Pay for Claude Seats and Only a Quarter of the Team Uses Them. How Do We Fix Adoption?
A recovery plan for a stalled Claude rollout: read the usage analytics you already have, sort people into three groups, give each a fix, and right-size seats instead of paying for empty ones.
Claude Can Join Your Team's Video Calls Now. What Should Admins Set Before It Does?
Claude Meetings is in beta on Team and Enterprise. What it does on a Google Meet call, which settings are on by default, and the five admin decisions to make before auto-join spreads.
How Do I Get a Non-Technical Team Actually Using Claude in the First 30 Days?
A week-by-week plan for getting a non-technical team using Claude chat and Cowork in 30 days: admin settings, one shared project, owner-provisioned skills, scheduled tasks and the analytics to check.
Claude Code Now Lets Slow MCP Tools Run in the Background
Claude Code 2.1.319 (November 6) adds support for MCP tasks in the terminal. A long-running tool call moves to the background and Claude gets the result when it finishes. Here's what that fixes, the two environment variables worth knowing, and what it won't do.
Claude Fable 5.2 Is Out. Should Your Team Switch, and Who Pays for It?
Claude Fable 5.2 shipped November 9, 2026. Which seats include it, why most teams should keep Opus 5.5 as the default, and the seven-day Enterprise switch admins need to check.
Claude Code Can Now Refuse to Write Outside Your Project, in Every Permission Mode
Claude Code v2.1.314 (October 30, 2026) added permissions.blockWritesOutsideWorkingDirectories, a setting that makes Write, Edit and NotebookEdit refuse any path outside your working directories. Here's what it covers, what it doesn't, and the settings file I'd put it in.
Teach Claude Code's Auto Mode About Your Setup in One Command
Claude Code 2.1.304 made /auto-mode-setup available in more places. It drafts the trusted-infrastructure list auto mode needs so it stops blocking your own repos. What it reads, what it writes, and the one mistake that wipes your safety rules.
Claude Code Now Compacts Your Session While You're at Lunch
Claude Code 2.1.299 added idle compaction: a session over 200K tokens that sits untouched for about 54 minutes gets summarized on its own. Here's why that interacts with the prompt cache, what it saves, and the one setting I'd check.
Claude Haiku 5.5 Is 20x Cheaper Than Sonnet. Here's Where I'd Actually Use It
Anthropic shipped Claude Haiku 5.5 on October 7, 2026: $0.10 per million input tokens, a five-level effort dial, and a 1M context window. It isn't your main coding model. It's the cheap, fast worker you hand the boring parts of your agents to. Here's the job split, the cost math, and the migration traps.
Claude Code Mods: Plugins That Can Change How Claude Code Itself Behaves
Claude Code v2.1.287 (Oct 1, 2026) added Claude Mods: plugins written in TypeScript that can block tool calls, rewrite prompts, and draw their own UI inside the CLI and desktop app. Here's what changed, what a mod looks like, and where I'd start.
Your Claude Code Default Just Changed to Opus 5.5. Here's What That Actually Costs You
Anthropic shipped Claude Opus 5.5 on September 22, 2026: cheaper and faster than Opus 5, matching Claude Fable 5.1 on most work. But it also quietly became the default model on the $20 Pro plan and Team Standard, replacing Sonnet. Here's what changed, what breaks, and what to check before it burns your usage window.
Claude Code Now Coordinates Its Own Parallel Sessions. I've Been Doing That Job By Hand.
Anthropic relaunched Claude Code Projects on September 17, 2026: one conversation that starts, tracks, and reports back on parallel cloud sessions ("threads") sharing memory, instructions, and repos. Here's what it actually replaces, and what it costs you in practice.
Claude Code Can Now Grade Your Own Skills (I Finally Have Proof Mine Work)
Anthropic shipped claude plugin eval in Claude Code v2.1.269 (Sep 11, 2026): a built-in way to score a plugin or skill against real prompts, compare it to a no-plugin baseline, and gate CI on the result. Here's how it works and what it caught in a skill I already run.
Claude Code Now Shows You the Diff While It's Still Editing
Claude Code shipped a live diff panel on September 3, 2026 (v2.1.260): a side-by-side view of every uncommitted change that updates as Claude edits, toggled with /diff. Here's what it actually shows, the Ctrl+X B trick most people miss, and why I stopped tailing git diff in a second terminal.
Claude Cowork Got Its Own Browser. The Isolation Model Is the Real Story
Anthropic shipped a built-in browser for Claude Cowork on August 26, 2026: no extension, no shared logins, one permission card per site. Here's what actually changed and why the access model matters more than the browsing itself.
Claude Academy Is Live: What Anthropic's Free Courses Actually Cover
Anthropic launched Claude Academy at academy.claude.com on August 20, 2026: free, certificate-bearing courses on Claude.ai, Claude Code, and the Claude Platform, built around a 4D fluency framework. Here's what's actually in it, and where a free self-paced course structurally stops short of what changes how you work.
Claude Code Stopped Asking Permission This Week
Auto mode became the default in Claude Code on August 14. Here's what the classifier actually checks before it acts without you, what it still refuses to do no matter what, and the two settings I changed before trusting it on real work.
Claude Code Sessions Can Talk to Each Other Now. Here's the Problem It Actually Solves.
Anthropic shipped cross-session messaging in Claude Code, and finished it this week: sessions can now find and message each other by name with an @-mention, no config required. Here's what it actually does, what it can't do, and how I'm using it.
Claude Merged Chat and Cowork Into One Project. The Mode Picker Is Gone
Anthropic quietly merged Chat and Cowork into a single Home on claude.ai web and desktop in early August 2026: one composer, and a project's memory, instructions, files, and scheduled tasks are now shared across both. Here's what changed and why the removed decision matters more than any new button.
MCP's Biggest Update Ever Shipped This Week. Most of It You'll Never Notice.
The Model Context Protocol had its largest spec revision since launch on July 28, 2026: a stateless core, hardened OAuth, and a new extension that lets tools render real interactive UI instead of text. Here's what changed, what breaks (nothing, yet), and the one part actually worth watching.
Claude Opus 5 Shipped. The Real Change Is the Effort Dial, Not the Benchmarks
Anthropic released Claude Opus 5 on July 24, 2026: frontier-class coding at half the price of Claude Fable 5, same pricing as Opus 4.8. But the change that will actually bite your code is a default-behavior swap around thinking and effort. Here's what shipped, what breaks, and what I'm doing differently.
Claude Code Artifacts Can Go Public Now. Live Dashboards Still Can't.
Anthropic shipped public sharing links, editor roles, and live MCP connector data for Claude Code Artifacts in the week of July 13, 2026. Here's what actually changed, and the one sharp trade-off that decides which feature you get.
Claude Cowork Is on Your Phone Now: What That Actually Unlocks
Claude Cowork just left the desktop. Anthropic shipped web and mobile access on July 7, and the real change isn't the new app icon, it's that a task can run all night without your laptop staying open. Here's what changed and who should care.
Claude Code's Default Model Just Changed: What Sonnet 5 Actually Gets You
Anthropic made Claude Sonnet 5 the default model in Claude Code on June 30, 2026, with a native 1M-token context window at standard pricing. Here's what actually changed, the benchmarks that matter, and what I'm doing differently this week.
Claude Code Can Now Ship Your Session as a Live Page
Claude Code Artifacts turn a coding session into a shareable, live-updating web page built from your session's full context. Here's what that means for solo builders and small teams.
What I Actually Use Claude Code For (Beyond Writing Code)
Claude Code writes my code, but that's maybe half of what it does for me. Here's how I run real business operations through it: Meta ad campaigns, Zoom scheduling, Stripe-driven inventory, and an SEO emergency, mostly in plain English over MCP. Including the parts that broke.
Claude Code Plugins: Inside Anthropic's Official Directory (2026 Guide)
Anthropic launched the official Claude Code plugin directory on May 22, 2026 - 55+ curated plugins covering language servers, MCP integrations, security review, git workflows, and more. Here's what's in it, how to install plugins, and the ones I'm actually running daily.
Best Claude Code Bootcamp in 2026: An Honest Comparison
An honest, current comparison of the ways to learn Claude Code in 2026: Anthropic's free Claude Academy and docs, DeepLearning.AI, self-paced courses, live Maven cohorts, private team training and 1:1 coaching.
How Long Does It Take to Get Good at Claude Code?
An honest, hour-by-hour and week-by-week timeline for getting from zero to productive to expert with Claude Code, from a working engineer who teaches it for a living.
Claude Code Can Watch Videos Now - Sort Of. Here's What Actually Shipped.
Claude Code didn't get a native video feature. But over the last two weeks, a cluster of plugins shipped that turn ffmpeg plus a transcript into a real video understanding loop. Here's the recipe, the trade-offs, and what it actually unlocks.
Claude Design: A First Look at Anthropic's Move Into the Visual Layer
Anthropic's new Claude Design tool turns prompts into prototypes, slides, and one-pagers - with a handoff bundle straight to Claude Code. Here's what it actually does, who it's for, and where the seams still show.
Claude Code Routines: What They Actually Do (And When They Beat a Cron Job)
Anthropic's new Routines feature turns Claude Code into a scheduled, cloud-run agent. Here's how it works, the daily limits per plan, and the workflows where it earns its keep over plain cron.
Claude Code Tutorial: The Complete Beginner's Guide
Everything a working engineer needs to go from zero to productive with Claude Code in an afternoon. Install, first session, permission model, CLAUDE.md, and five things to try in your first ten minutes.
Claude Code vs Cursor vs GitHub Copilot: An Honest Comparison
A working engineer's honest comparison of the three most popular AI coding tools. What each does well, where each falls short, cost breakdown, and my actual setup.
How to Write a CLAUDE.md File (With Examples)
What to put in a CLAUDE.md file, what to leave out, and two complete examples you can copy: a Next.js app and a Python API.
What Is the Claude Agent SDK? A Developer's Guide
The Claude Agent SDK is Claude Code's engine, programmable in Python and TypeScript. A working engineer's guide to what it does, when to use it, and a complete hello-world agent you can run today.
Claude for Enterprise: How Engineering Teams Are Using Claude in Production
The problem with ad-hoc Claude adoption at work - and what the teams who are getting it right actually do. Shared CLAUDE.md, team-wide commands, MCP servers, and why training is the gap nobody's filling.