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

# Playground

> Explore your codebase and knowledge base interactively through a free-form AI chat interface.

The **Playground** is a multimodal chat interface powered by the same AI agent that resolves support tickets. Use it to explore your codebase, surface knowledge base articles, or troubleshoot issues in a free-form conversation.

## How to Use the Playground

<Steps>
  <Step title="Set your context">
    Add or remove **repositories** to control which codebases the agent searches. Use **@ mentions** to pin a specific ticket or knowledge base article as context.
  </Step>

  <Step title="Ask a question">
    Type your question in the input box and press **Enter**. Attach files like screenshots, log files, PDFs, or recordings to give the agent the full picture.
  </Step>

  <Step title="Continue the conversation">
    Follow up with additional questions. Each message builds on the full history of the conversation.
  </Step>
</Steps>

## Working with Answers

Once the agent completes its response, you can:

* **Sources** — View every source the agent used (knowledge base articles, code files, logs), grouped by type
* **View Reasoning** — See the agent's step-by-step investigation process
* **Copy** — Copy the full response to your clipboard
* **Create Article** — Turn a useful answer into a knowledge base article with **In Review** status

<Tip>
  The Playground is a great tool for onboarding new team members. They can ask questions about the codebase and product without needing to find the right person to ask.
</Tip>

## Conversation History

All conversations are automatically saved. Access past conversations by clicking **History** in the context bar or pressing **⌘K** (Mac) / **Ctrl+K** (Windows/Linux).

## Providing Feedback

After each response, click **Yes** or **No** under the "Was this helpful?" prompt and optionally leave a comment. Feedback helps the Decimal team improve the agent over time.