> ## Documentation Index
> Fetch the complete documentation index at: https://autter.dev/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Context

> Use attribution and prompt records to understand why AI-authored code exists.

Autter gives you the context behind AI-authored code without guessing from the code alone.

When you need to understand a change, start with the artifacts Autter already records:

* Line-level attribution in Git notes
* Commit-level statistics
* Prompt records when they are available in the active storage backend
* Session and model identifiers for the coding agent that made the change

## What context can show you

| Question                                      | Use                              |
| --------------------------------------------- | -------------------------------- |
| Who wrote this line?                          | `autter blame <file>`            |
| How much of this commit came from AI?         | `autter stats <commit-or-range>` |
| What prompt or session created this change?   | `autter show-prompt <id>`        |
| What does the current working tree look like? | `autter status`                  |

<Note>
  `autter show-prompt` only returns prompt records that exist in the current storage backend. In local-only mode, that usually means local data on the machine.
</Note>

## Use context when reviewing code

When a line looks surprising, use the surrounding authorship data first:

1. Check the file with `autter blame`.
2. Open the relevant commit with `autter stats` or `autter show`.
3. Read the prompt record with `autter show-prompt` if the backend has one.
4. Compare the result with the current repository history.

This gives you the intent behind the code, not just the final diff.

## Keep context available to your team

Connected mode can make prompt and attribution data available to the Autter platform for shared analytics and audit history.

Local-only mode keeps the same line-level attribution on the machine and in Git notes. Use it when prompts must stay off the platform.

<CardGroup cols={2}>
  <Card title="AI blame" icon="code-branch" href="/cli/ai-blame">
    Inspect line-level authorship in a file.
  </Card>

  <Card title="Data and privacy" icon="shield" href="/cli/data-and-privacy">
    See what stays local and what can sync when you connect.
  </Card>
</CardGroup>