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

# Overview

> Give Prompt to Video real knowledge to draw from — knowledge bases, MCP servers, and connected applications.

Connectors are how Prompt to Video learns about your product, your data, and your team's context before it writes a script. Without a connector, the agent works from your prompt alone. With one, it can pull in real, current information instead of guessing.

<Frame>
  <img src="https://mintcdn.com/velo/ges-rbAcLDEpYx4O/images/prompt-to-video/connectors/01.png?fit=max&auto=format&n=ges-rbAcLDEpYx4O&q=85&s=4a58875fb0ca17a73c975ee025bfbfb6" alt="Connectors tabs" width="1468" height="837" data-path="images/prompt-to-video/connectors/01.png" />
</Frame>

Connectors live under **Connectors** in your workspace sidebar, split into three tabs:

## Knowledge bases

Turn a website or set of docs into a reusable, queryable knowledge source. Crawl a site once, select the pages that matter, and the agent can reference that content across every future video. See [Knowledge bases](/prompt-to-video/connectors/knowledge-bases).

## MCP servers

Bring your own MCP (Model Context Protocol) server as a knowledge source, for teams with internal tools or data that aren't covered by a standard integration. See [MCP servers](/prompt-to-video/connectors/mcp-servers).

## Applications

Connect over 100 apps — Slack, Notion, Gmail, Google Drive, Confluence, Salesforce, HubSpot, Linear, Jira, Zendesk, GitHub, Google Calendar, and more — so the agent can pull relevant context directly from the tools your team already uses. See [Applications](/prompt-to-video/connectors/applications).

## Do I need a connector?

No — Prompt to Video works from a prompt alone. Connectors are for when you want the script grounded in specific, current information rather than general knowledge.