Skip to main content
The Model Context Protocol (MCP) lets an AI assistant call tools in another application. Aspect’s hosted MCP server gives your assistant access to the media and projects you can use in Aspect. Connect this Streamable HTTP endpoint:
Use it to find shots, read transcripts and document text, organize selects, work with canvases, and manage review. The tool reference lists the available calls.

Connect your assistant

In Claude, open Settings → Connectors, add a custom connector, and enter https://api.aspect.inc/mcp. Connect it and sign in to Aspect when prompted. Your plan or workspace must allow custom connectors.This is a remote connector. It is separate from a local desktop extension such as the DaVinci Resolve MCP.

Authentication and permissions

OAuth opens an Aspect sign-in flow and is the recommended option for clients that support it. For a client that accepts explicit headers, you can use an Aspect API key as an Authorization: Bearer token instead. Create a key using the API overview and keep it in your client’s secret or environment configuration. Every tool call follows the signed-in user’s permissions. A connection does not grant access to another workspace or override a project’s restrictions. The local CLI has a separate authentication state; check aspect auth status --json before combining it with MCP.
Aspect also provides connection instructions in Settings → Aspect Agent & MCP. Start with a read-only request: “List the Aspect workspaces and projects I can access.”

Choose the right tool for the work

Aspect is not a video editor. It provides the media, context, and file access that let an agent work effectively in an editor.

Search and organize media

Turn a brief into a traceable list of clips and time ranges.

Agentic Video Editing

Edit from mounted proxies while camera originals pin in the background.

Tool reference

Browse the available calls and their boundaries.

Troubleshooting

Check connections, permissions, missing results, and background work.