samuelgursky/davinci-resolve-mcp

What it solves

This project provides a bridge between AI assistants and DaVinci Resolve Studio, allowing LLMs to programmatically control the video editing software. It eliminates the manual effort of repetitive editing tasks, media organization, and project management by exposing the official Scripting API to AI agents.

How it works

The project implements a Model Context Protocol (MCP) server. It can connect to DaVinci Resolve in two ways: via the official external scripting API (for Studio users) or through an in-app bridge script for the free edition.

It offers two server modes:

• Compound Mode: Groups related operations into a few high-level tools to save context window space.
• Full/Granular Mode: Provides a one-to-one mapping of every Resolve API method as an individual MCP tool.

Additionally, an Advanced Server allows for offline editing of Resolve project files (.drp, .drt, .drx) and database changes without needing the software running.

Who it’s for

Video editors, colorists, and post-production houses who want to automate their workflows using AI agents (such as Claude, Cursor, or VS Code) to handle ingest, sync, organization, and first-pass assemblies.

Highlights

• Full API Coverage: Supports 100% of the official DaVinci Resolve Scripting API.
• Multi-Edition Support: Works with both the Studio and Free editions of Resolve.
• Cross-Platform: Validated on macOS, Windows, and Linux.
• Source-Safe Analysis: Includes tools for media analysis (transcription, 2-pop detection) that read source files without modifying them.
• Offline Project Manipulation: Ability to edit project files and database entries without Resolve open.
• Local Control Panel: A browser-based interface for inspecting state and managing analysis jobs.