DavinciVideoMcp
Provides integration with Nextcloud for managing a creative asset library, including cataloging, retrieval, and local caching.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@DavinciVideoMcpAnalyze the current clip for scene cuts and add markers"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
DavinciMcp 文档基线
本目录是新工程的产品、架构、实施与专业 Skills 基线。
当前开发状态(重要)
当前 Web 仅是开发测试工作台,不是正式自动剪辑交付界面;
videos/中三条素材只用于验证受管 Resolve 写入、读回和渲染,输出只能称为内部技术预览;内部技术预览、内部工作版和用户可见的成片候选是三类不同产物。技术预览不包含转写、多模态理解、素材理解、故事选片、字幕、音乐、音效、降噪、运镜、动效或收尾,绝不能标记为
candidate-v1或ready_for_review;正式成片候选必须有完整媒体证据、素材理解、EditPlan、CapabilityBinding、工作版复核、收尾方案和候选验证;缺少任一项时,任务会安全停在等待条件状态,不会调用 Resolve 继续“凑成片”。Gemini 直连音视频不可用本身不是停机条件:首版改用 FunASR、本地声音测量、按需密集抽帧和 Codex 图片分析,并披露其边界;
当前创意库的认证能力数为
0。未经“发现、部署、执行、读回、渲染”认证的音乐、字体、模板、动效和其他创意资源均不可使用,也不可声明支持。
Related MCP server: resolve-mcp
阅读顺序
AGENTS.md:开发原则、语言要求、范围控制和完成后汇报;docs/PRODUCT.md:产品目标、用户流程、范围和验收;docs/ARCHITECTURE.md:模块、进程、数据、任务和扩展边界;docs/CREATIVE_LIBRARY.md:Nextcloud 创意素材库、目录、检索与本地缓存;docs/MEDIA_INTELLIGENCE.md:上传校验、FunASR、多模态证据和素材理解实现;docs/DAVINCI_ENGINE_MCP.md:自研 Resolve 执行 MCP、参考仓库、能力迁移与验收合同;docs/CREATIVE_ADAPTER_DEVELOPMENT.md:创意能力 Adapter 的当前实现、认证边界和新增方式;.agents/skills/*/SKILL.md:各专业任务的判断方法与交接;docs/IMPLEMENTATION_PLAN.md:本次开发的纵向切片;
权威范围
开发过程与交付说明以
AGENTS.md为准;产品需求以
docs/PRODUCT.md为准;系统实现边界以
docs/ARCHITECTURE.md为准;创意素材管理与检索以
docs/CREATIVE_LIBRARY.md为准;媒体分析与素材理解输入以
docs/MEDIA_INTELLIGENCE.md为准;Resolve 执行合同、外部实现参考和能力迁移边界以
docs/DAVINCI_ENGINE_MCP.md为准;专业判断方法以对应
SKILL.md为准;精确字段、枚举和接口参数以代码 Schema 与自动生成接口文档为准。
文档之间出现冲突时,不自行折中:先按以上权威范围定位责任,再向产品负责人确认。
本机配置
.env.example:多模态反代、Nextcloud、缓存和工作区配置示例;数据库、项目文件和创意缓存默认统一放在./workspace/;models/manifest.example.yaml:FunASR 本地模型配置示例;
真实密钥、模型权重、数据库、工作区和本地缓存不得提交 Git。
固定运行环境
本项目首期复用现有 Conda 环境,不得自动创建新的 Python 环境:
Conda 环境名:unofficial-davinci-mcp-win
Python 版本:3.10.20开发、安装依赖和启动前使用:
conda activate unofficial-davinci-mcp-win
python --version
where pythonAPI、Worker 和 davinci-engine-mcp 必须使用同一个解释器。未经产品负责人明确同意,不得执行 conda create、python -m venv、uv venv、Poetry 自动建环境或创建 .venv。项目采用 src 目录布局,scripts/start.ps1 会先设置必要的 PYTHONPATH 再调用统一启动器;手动执行模块命令时也应先设置该路径:
$env:PYTHONPATH = "$PWD\src;$PWD\davinci-engine-mcp\src"
conda run --no-capture-output -n unofficial-davinci-mcp-win python -m davinci_app启动时必须校验 CONDA_DEFAULT_ENV、Python 版本和 sys.executable;不符合时直接停止并提示激活现有环境,不能自行创建替代环境。
本机启动
在仓库根目录执行:
.\scripts\start.ps1该脚本只复用 unofficial-davinci-mcp-win,启动 API 与持久 Worker;Engine MCP 仅由 Worker 按需以 stdio 子进程启动。
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- Flicense-qualityBmaintenanceA Model Context Protocol server that enables AI assistants like Claude to interact with DaVinci Resolve Studio, providing advanced control over editing, color grading, audio, and other video production tasks.Last updated17
- Alicense-qualityFmaintenanceComprehensive MCP server for DaVinci Resolve with 295+ tools to control projects, timelines, editing, color grading, rendering, and more via natural language.Last updated3MIT
- AlicenseAqualityBmaintenanceAn MCP server that lets AI agents visually perceive and edit video inside DaVinci Resolve Studio by rendering frames and manipulating the timeline.Last updated2MIT
- Alicense-qualityDmaintenanceA Model Context Protocol (MCP) server that connects AI coding assistants to DaVinci Resolve, enabling query and control through natural language.Last updatedMIT
Related MCP Connectors
MCP server for Google Veo AI video generation
MCP server for Wan AI video generation
MCP server for Luma Dream Machine AI video generation
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/angelOnly/DavinciVideoMcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server