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

# Knowledge Base

> Upload and organize documents for AI workers and Radbot to access

The Knowledge Base is your central repository for documents that AI workers and Radbot can access to provide more accurate and contextual responses. Upload documents into organized collections to enhance your AI-powered security workflows.

## How It Works

Upload documents into **Collections** that AI workers and Radbot can reference when performing tasks. This gives your AI agents access to your organization's specific knowledge, policies, and historical data.

<Note>
  **Collections organize your documents** - Group related documents together so AI workers can find relevant information quickly and provide more accurate responses.
</Note>

## Document Examples

<AccordionGroup>
  <Accordion title="Incident Response">
    Incident reports, playbooks, post-mortems, and threat intelligence (IOCs, TTPs)
  </Accordion>

  <Accordion title="Compliance & Governance">
    Security policies, compliance frameworks (SOC 2, ISO 27001), audit reports, risk assessments
  </Accordion>

  <Accordion title="Technical Documentation">
    Network diagrams, asset inventories, vulnerability reports, configuration baselines
  </Accordion>

  <Accordion title="Threat Intelligence">
    IOC feeds, threat reports, vulnerability databases, security advisories
  </Accordion>
</AccordionGroup>

## Getting Started

1. **Create Collections** - Organize documents by purpose (incidents, compliance, technical, etc.)
2. **Upload Documents** - Add PDFs, Word docs, text files, and other supported formats
3. **Tag Documents** - Add metadata to help AI workers find relevant information
4. **Test with Radbot** - Ask questions to verify your documents are accessible

<Tip>
  **Pro Tip:** Start with your most frequently referenced documents like incident playbooks and security policies. These will have the biggest impact on AI worker performance.
</Tip>
