Connect VAUD.
Give your AI agent access to the editing library.
Add the connection.
Use an MCP client that supports Streamable HTTP. Add a remote server named VAUD and choose no authentication.
For Claude Code:
claude mcp add --transport http vaud https://mcp.vaud.studio/mcp
For URL-based JSON configurations:
{"mcpServers":{"vaud":{"url":"https://mcp.vaud.studio/mcp"}}}Start with a look.
Ask: "Use VAUD to help me choose an editing style. Read the start and onboarding guides first."
Your agent can search 262 effects, read their editable source and fetch the production guides and scripts. Shared runtime assets and fonts are hosted here too. The connection works without GitHub access.
Your agent uses its own tools to edit and render your footage. This public library connection has no access to private customer accounts or publishing services.
A browser visit to the MCP URL returns 405 because the endpoint accepts MCP POST requests. Use the URL in your MCP client's settings.