> For the complete documentation index, see [llms.txt](https://www.pranaypourkar.co.in/the-programmers-guide/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://www.pranaypourkar.co.in/the-programmers-guide/ai/claude/claude-fundamentals/what-is-claude.md).

# What Is Claude?

## About

Claude is an AI system made by Anthropic. We can use it to read, write, reason through problems, work with code, and understand images or documents.

At its simplest, Claude is a conversation partner: we give it a question or some material, and it responds. But it becomes much more useful when we give it a real task, the right context, and a clear definition of a good result.

## A Simple Example

Instead of asking:

> Explain this error.

You can ask:

> Read this error and the related method. Tell me the likely cause, show the smallest safe fix, and suggest one test that would catch it next time.

The second request gives Claude a job, a boundary, and a useful outcome to aim for.

## What Claude Can Help With

Claude works well when there is information to understand or transform.

* **Writing and analysis** — summarising a document, comparing options, drafting an explanation, or turning notes into a plan.
* **Software work** — reading code, debugging, suggesting changes, writing tests, and reviewing a pull request.
* **Documents and images** — extracting information from PDFs, screenshots, diagrams, and other visual inputs.
* **Structured outputs** — producing a checklist, table, JSON response, or another defined format.
* **Tool-based work** — using connected tools or services when they are available and permission has been given.

{% hint style="info" %}
The quality of the answer depends heavily on the context we provide. A useful prompt is not necessarily long; it is specific about the task, the relevant information, and what “done” looks like.
{% endhint %}

## Claude Is Not Magic

Claude predicts and reasons from the information it receives. It can make mistakes, miss important context, or sound confident when an answer needs checking.

For anything that matters, keep a human in the loop:

* Check facts against the original source.
* Run the tests before accepting code changes.
* Review security-sensitive or production-facing work.
* Ask Claude to explain its assumptions when the answer is unclear.

This is not a weakness unique to Claude. It is simply the right way to work with AI: use it to move faster, then apply engineering judgment.

## Claude Is More Than One Interface

We may meet Claude in several places:

| Where we use it   | What it is useful for                                      |
| ----------------- | ---------------------------------------------------------- |
| **Claude.ai**     | Chatting, research, analysis, writing, and artifacts       |
| **Claude Code**   | Working directly in a codebase from the terminal or editor |
| **Claude Cowork** | Completing file-based tasks and supervised workflows       |
| **Claude API**    | Adding Claude to your own product or automation            |

They are different ways of using Claude, not unrelated products. The same ideas- clear instructions, useful context, tool access, and verification apply everywhere.

## Where to Go Next

If you are new to Claude, read **Claude Models** next to understand the different capability levels. If you already have a task in mind, jump to **Working with Claude** for everyday work or **Claude Code** for software development.

## Official Documentation

* [Anthropic: Intro to Claude](https://platform.claude.com/docs/en/intro)
* [Anthropic: Features overview](https://platform.claude.com/docs/en/build-with-claude/features-overview)


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://www.pranaypourkar.co.in/the-programmers-guide/ai/claude/claude-fundamentals/what-is-claude.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.