Claude AI Resource Hub

For new developers

If you have never used an AI coding assistant before, the shortest path to a working session is the getting started guide. It assumes you can open a terminal and that you have a package manager installed — nothing beyond that. The guide walks through installing the CLI, setting the required environment variable, opening a project, and running the first prompt, with each step's expected output shown so you know whether something worked or silently failed.

Once the CLI is running, the free tier summary is worth reading before you start a long session, because understanding the daily usage cap saves the frustration of hitting it mid-task without knowing why. The claude code overview then explains what the tool can and cannot do — reading that before writing your first complex prompt tends to produce better outcomes than discovering the limits by bumping into them.

For infrastructure engineers

Infrastructure engineers typically arrive with a specific question: how do we deploy this safely, and what does it cost at scale? The claude api reference is the right starting point for the HTTP surface — it covers endpoints, authentication, rate limits, and the data-handling parameters that matter for a production integration. The trust and safety page follows naturally: it covers .claudeignore configuration, least-privilege agent permissions, and the specific CI/CD considerations that come up when you are running the tool in an automated pipeline rather than interactively.

For teams managing a larger rollout, the enterprise configuration page covers audit log integration with Splunk and Datadog, OIDC identity provider setup, and the flags that bound agent autonomy in regulated environments. The models overview rounds out this path with context window and throughput data that is relevant when designing prompt templates for high-volume pipelines.

For team leads

Team leads evaluating Claude AI for their squad face three questions in sequence: does it work, is it safe to use with our codebase, and what will it cost? The for teams reference addresses the first by covering shared configuration, the skills registry, and how to keep an entire squad on the same CLI version and flag set without a bespoke onboarding runbook per engineer. The trust and safety page addresses the second with the data handling and access control notes needed for a policy conversation. The free tier summary and models overview together address the third.

Team leads who have already cleared the policy hurdle and are planning the rollout usually benefit from reading the skills reference next, because skills are the mechanism by which team-specific workflows — release checklists, migration helpers, incident runbooks — get packaged as reusable CLI extensions rather than living in someone's personal dotfiles.

For curious non-coders

If you are not a developer but want to understand what Claude AI is and what the toolchain does, the about this reference page explains the scope of the hub in plain language. The getting started guide is written to require no prior command-line experience and each step is explained rather than assumed. The models overview is the most accessible technical page — it explains the difference between Opus, Sonnet, and Haiku in practical terms without requiring you to understand the underlying architecture.

For context on the broader AI safety conversation, the Stanford HAI initiative publishes accessible explainers aimed at a general audience. The NIST AI Risk Management Framework is a government resource that covers AI risks and governance in terms designed for policy readers, not just engineers.

Topic	Best starting page	Related pages
First install, any platform	Getting started	Install claude code, Windows
Model selection and pricing context	Models overview	Free tier, API pricing
Data privacy and access control	Trust & safety	Enterprise, API reference
Team rollout and shared config	For teams	Skills, Enterprise
CLI extension and automation	Skills reference	API reference, For teams
Plain-language AI overview	About this reference	Getting started, Models overview

How to use this hub

Every page in this reference is designed to be read independently. You do not need to start at the home page and work through sections in order. The breadcrumbs at the top of each page show where you are in the hierarchy; the related topics section at the bottom of each page lists the most natural next steps from that particular starting point. If you land on a page that is not the right level of detail — either too technical or not technical enough — the related-block links at the bottom will point you to adjacent pages that cover the same topic from a different angle.

Pages are reviewed on the cadence listed on the about page. When a page has been recently updated after a CLI release, the editors note the change at the top of the relevant section. If something looks wrong or out of date, the contact page is the fastest route to a correction.