The complete curriculum for Claude — Projects, Artifacts, computer use, Claude Code, and the conversation patterns that get the best output from the most thoughtful AI on the market.
What Claude can do beyond chat: Cowork, Claude Code, Computer Use, MCP. Real examples of each agent and what becomes possible.
Start here — real value before you ever upgrade.
Set up Microsoft 365 integration with Claude. Then 5 use cases: triage, drafting replies, finding important threads, weekly summaries, email-to-task automation.
Anthropic replaced its Slack app with a persistent @Claude that lives in a channel, remembers context, and can be handed work by anyone on the team. What it does, how it differs from the old app, and how to use it without the noise.
Launched June 9, switched off three days later by a US export-control order — then reinstated July 1 with a new safety classifier. The full timeline, what changed on the way back, and where Fable 5 fits the lineup now.
Anthropic's June 2026 Mythos-class release in plain English: the capability jump, why there are two names (Fable vs Mythos), the safeguards, and its June suspension and July 2026 reinstatement.
Launched July 9 in beta: a Settings dashboard showing what you use Claude for, when, and how well you collaborate — plus quiet hours, break nudges, and the privacy fine print worth reading.
On June 30, 2026 Anthropic made Sonnet 5 the default for Free and Pro users — near-Opus performance at a fraction of the cost, with a 1M-token window. What changed, what it costs, and when to still reach for Opus 4.8 or Fable 5.
Free, Pro, Max — plus the API tiers (Haiku, Sonnet, Opus). The model-choice lever that quietly controls your bill.
The everyday power-user workflows — included in both paid tiers.
Project setup, system instructions, custom knowledge base, conversation memory. Build a research assistant that knows your industry, your style, and your priorities.
When to invoke Artifacts, when to skip them, and the iterative-refinement workflow that turns Claude into a coding/writing partner instead of just an answer machine.
Loading entire documents, multi-document comparison, research synthesis. The use cases where Claude's context window genuinely changes what's possible.
Setting up Claude's computer-use feature, the workflows worth automating (and the ones that aren't yet), safety and oversight patterns.
How to write custom instructions that genuinely change output. The 4 sections every good instruction set needs, and the anti-patterns that make Claude worse.
Cowork turns Claude into a real working environment — desktop, and now the cloud (web + mobile rolling out July 2026). Claude reads your files, runs scripts, drafts documents, automates the boring stuff.
Most people don't realize how much time they lose to file shuffling — renaming, sorting, converting, extracting.
Cowork's biggest practical win is creating Office-quality documents from scratch.
The most powerful model isn't always right. A 3-question decision framework by task, speed, and cost — and how to spend usage credits wisely after the June 23 pricing change.
The deep, power-user lessons — exclusive to Pro+.
For technical roles. Setting up Claude Code, prompt patterns that actually ship code, MCP servers, hooks. The tool that turned vibe-coding into a real workflow.
Multi-turn agent workflows, connecting Claude to external tools via MCP, prompt chaining, evaluation patterns. The advanced curriculum for power users.
Email is where most knowledge workers lose 1-2 hours a day. With Cowork connected to Outlook (or Microsoft 365), the morning triage that used to take 90 minutes becomes a structured 15-minute review of what actually needs your attention.
Out-of-the-box Cowork is powerful. But its real superpower is extensibility — three different mechanisms for adding capabilities.
The last Cowork capability is the one that changes the relationship: scheduled tasks and live artifacts.
When a model works autonomously for hours, the skill shifts from clever prompts to good briefs. Four delegation patterns: goals, constraints, checkpoints, and self-validation.
Why some requests get answered by Opus 4.8, why harmless ones occasionally trip the filter, and what the new 30-day data retention means for your company's AI policy.
All 23 Claude lessons are available today — plus every other tool track, the hands-on builds, and the playground. Founding pricing — $9/mo Pro, $19/mo Pro+ — locks your rate for life: the price you join at never goes up.
Start free →