trident-mcp

trident-mcp

AI 3D model generation and post-processing MCP server — text/image/multiview-to-3D via Tripo, retopology, format conversion (GLB/FBX/OBJ/STL/USDZ), and stylization. Single Go binary, 10 tools.

Category
访问服务器

README

trident-mcp

Go License trident-mcp MCP server

<p align="center"> <img src="assets/pipeline-showcase.png" alt="Pipeline showcase: reference images → AI concept → multi-view generation → 3D model" width="100%"> </p>

trident-mcp is a Go MCP server for AI-assisted 3D model generation and post-processing.

trident-mcp MCP server

The server is client-agnostic and works independently with any MCP-compatible client. You do not need any companion skills or extra MCP servers to use the core 3D generation, polling, download, and post-processing tools.

It currently ships with a Tripo-backed provider and exposes tools for:

  • text-to-3D generation
  • image-to-3D generation
  • multiview-to-3D generation
  • async task polling
  • model download
  • retopology
  • format conversion
  • stylization
  • model catalog and server config inspection

Requirements

  • Go 1.25+
  • A Tripo API key in TRIPO_API_KEY

Install

Build locally:

go build ./cmd/trident-mcp

Or install with Go:

go install github.com/mordor-forge/trident-mcp/cmd/trident-mcp@latest

Configuration

The server reads configuration from environment variables:

Variable Required Default Description
TRIPO_API_KEY Yes none Tripo API key used for generation and editing calls
MODEL_OUTPUT_DIR No ~/generated_models Directory where downloaded models are written

Running

The server speaks MCP over stdio:

TRIPO_API_KEY=tsk_your_key_here trident-mcp

Example MCP client configuration:

{
  "mcpServers": {
    "trident-mcp": {
      "command": "trident-mcp",
      "env": {
        "TRIPO_API_KEY": "tsk_your_key_here",
        "MODEL_OUTPUT_DIR": "/absolute/path/to/generated_models"
      }
    }
  }
}

Tools

Generation

  • text_to_3d
  • image_to_3d
  • multiview_to_3d

These tools start asynchronous tasks. Use task_status to poll for completion, then download_model to retrieve the task output.

Status and Download

  • task_status
  • download_model

download_model saves the task's actual output format. If you need a different format, run convert_format first and then download the conversion task.

Post-processing

  • retopologize
  • convert_format
  • stylize

Introspection

  • list_models
  • get_config

get_config reports the active backend, output directory, and server version.

Skills

The repo also includes companion agent skills under skills/:

  • skills/3d-gen/SKILL.md
  • skills/multiview-3d/SKILL.md
  • skills/3d-to-blender/SKILL.md

These skills are optional. The MCP server itself works fine on its own in any MCP client.

Some of the companion skills are designed to compose trident-mcp with gemini-media-mcp for a fuller automated pipeline. In that setup, gemini-media-mcp can help with ideation, reference image generation, and multi-angle image creation, while trident-mcp handles reconstruction and post-processing. That pairing enables a more complete flow from idea to finished 3D model.

Development

Run the local checks:

go test ./...
go vet ./...

There is also an opt-in E2E smoke test for live Tripo uploads:

TRIPO_API_KEY=tsk_your_key_here go test -tags=e2e -run "TestE2E_" ./internal/provider/tripo/ -v

Release

GitHub Actions runs unit checks on pushes and pull requests, and runs the E2E smoke test on main and version tags. Tagged releases are built with GoReleaser.

Artifacts are stamped with the release version so the binary and MCP implementation metadata stay aligned.

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选