Full-coverage Model Context Protocol (MCP) server for DaVinci Resolve Studio — control editing, media, color, Fusion, Fairlight, AI features, and rendering from any MCP client.
12 tools with consistent naming (verb_noun pattern: open_page, get_current_page, list_layout_presets, etc.). All have descriptions (avg ~80 chars, within baseline 34-392). Input schemas present for all tools with proper type declarations. However, descriptions are terse and lack context for LLM selection, most are single-sentence statements without WHEN/WHY guidance. Parameter descriptions are minimal (e.g., 'name of a previously saved layout preset' lacks format/constraint details). No output schemas documented. Error handling and recovery guidance absent. Composition is sound (single responsibility per tool), but descriptions fall short of LLM-optimized baseline (50-200 chars).
Delete a saved UI layout preset by name.
Export a saved UI layout preset to a file on disk.
Get the currently active page in DaVinci Resolve (Media/Cut/Edit/Fusion/Color/Fairlight/Deliver).
Get the current Color page keyframe mode (0=All, 1=Color, 2=Sizing).
Get the DaVinci Resolve product name and version string as JSON.
Get the DaVinci Resolve version as structured fields (major/minor/patch/build/suffix) JSON.
Import a UI layout preset from a file on disk.
Descriptions lack LLM-optimized context. Most are single-sentence statements (avg ~80 chars) without WHEN to use, WHAT it returns, or dependencies. Baseline for A-grade tools is 50-200 chars with explicit guidance.
No output schemas documented. LLMs cannot plan downstream calls or extract required fields. E.g., get_version returns 'structured fields (major/minor/patch/build/suffix) JSON' but the exact schema is not visible.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 59 | <=2025-11-25 | v2 |
List saved DaVinci Resolve UI layout presets found on disk, as JSON.
Load a saved UI layout preset by name.
Switch DaVinci Resolve to a specific page.
Save the current UI layout as a new named preset.
Overwrite an existing UI layout preset with the current layout.
Parameter descriptions are minimal and lack format/constraint details. E.g., 'preset_name' is described as 'name of a previously saved layout preset' but does not specify allowed characters, length limits, or case sensitivity.
No error handling or recovery guidance. Destructive tools (delete_layout_preset) lack confirmation or dry-run patterns. No guidance on what to do if a preset is not found or if an operation fails.
Missing dependency documentation. E.g., open_page accepts 'media', 'cut', 'edit', etc., but no description explains what these pages are or when to use each. LLMs cannot reason about page selection without context.