PluginBench
MCP Server
Maintained

io.github.AdonaiVera/fiftyone-mcp-server MCP Server

io.github.AdonaiVera/fiftyone-mcp-server

Control FiftyOne computer vision datasets through AI assistants with 45+ MCP tools for data operations and App UI control.

RepositoryRegistry
How to choose

What this kind of tool is for

  • Query, inspect, or seed a database without writing a one-off script.
  • Let your agent look up schema details before it writes migration code.
  • Move a small dataset between tools without leaving the terminal.

How to compare servers like this

Engine match:
Confirm it speaks your exact database and version, not just the family.
Access scope:
Read-only connections are the safe default — start there before granting writes.
Safety rails:
Check whether destructive statements need an explicit confirmation before they run.
Maintenance signal:
Database drivers change; a recently updated tool is less likely to break mid-query.

This page carries your copy-paste config: pick your MCP client in the install section on this page and use the copy button.

What is the io.github.AdonaiVera/fiftyone-mcp-server MCP server?

The FiftyOne MCP Server enables AI assistants to explore, analyze, and manipulate computer vision datasets through natural language. It exposes 45+ tools across dataset management, App UI control, operators, plugins, aggregations, and schema operations. The server integrates with Claude, Cursor, and other AI tools to automate FiftyOne workflows.

This MCP server bridges FiftyOne's powerful computer vision dataset platform with AI assistants. You can query datasets, execute operators, control the FiftyOne App UI in real time, discover and install plugins, run aggregations, and manage samples—all through natural language commands. It's designed for both interactive App-based workflows and headless terminal environments.

How to install io.github.AdonaiVera/fiftyone-mcp-server

Copy-paste configuration for popular MCP clients.

transport: stdio
Config generated by PluginBench — verify against the source before use.
~/Library/Application Support/Claude/claude_desktop_config.json
JSON config
{
  "mcpServers": {
    "fiftyone-mcp-server": {
      "command": "uvx",
      "args": [
        "fiftyone-mcp-server"
      ]
    }
  }
}

Tools & capabilities

Tools this server exposes to the agent.

  • Dataset Management — List, load, and summarize datasets
  • App Operations — Control the App UI including views, panels, selection, and 25+ additional controls
  • Operator System — Discover and execute any FiftyOne operator
  • Pipelines — Run pipelines and manage delegated operations
  • Plugin Management — Discover, install, and manage plugins
  • Session — Launch the FiftyOne App server
  • Aggregations — Count, distinct, bounds, mean, histogram, and other aggregation operations
  • Samples — Add, tag, untag, and set values on samples
  • Schema — Inspect and modify dataset field schemas
  • App Config — Configure color scheme, sidebar groups, and active fields

Use cases

  • List and load computer vision datasets, then generate summaries and statistics through natural language queries
  • Find similar or near-duplicate images in datasets using AI-driven search and filtering
  • Control the FiftyOne App UI in real time—open panels, set views, select samples, and manage visualizations without manual interaction
  • Discover and automatically install FiftyOne plugins (brain, zoo, annotation, evaluation) when needed for specific tasks
  • Execute custom operators and pipelines on datasets to automate data processing workflows

io.github.AdonaiVera/fiftyone-mcp-server MCP server FAQ

What is the FiftyOne MCP Server?

It's an MCP server that gives AI assistants like Claude and Cursor access to FiftyOne's computer vision dataset platform. You can query datasets, execute operators, control the App UI, manage plugins, and run aggregations through natural language.

Is it free?

Yes. The server is open-source under the Apache 2.0 License and available on PyPI. FiftyOne itself is also open-source.

How do I install it in Claude or Cursor?

Install via pip: `pip install fiftyone-mcp-server`. Then configure your AI tool by adding the server to its MCP config (e.g., `claude mcp add fiftyone -- fiftyone-mcp` for Claude Code, or edit the config JSON for Claude Desktop/Cursor).

Do I need authentication or API keys?

No. The server works with local FiftyOne datasets and the App. No external authentication is required.

What are the different tool modes?

SDK mode works everywhere (datasets, aggregations, operators, plugins). APP mode controls the FiftyOne App UI in real time (requires a connected browser). SESSION mode launches a local App server from the terminal.

Can the AI install plugins automatically?

Yes. The server exposes plugin discovery and installation tools, and the AI can discover and install plugins (brain, zoo, annotation, evaluation, etc.) when needed for tasks.

README (reference)

Source of truth, from the repository.

FiftyOne MCP Server

<!-- mcp-name: io.github.voxel51/fiftyone-mcp-server --> <div align="center"> <p align="center"> <!-- prettier-ignore -->

<img src="https://user-images.githubusercontent.com/25985824/106288517-2422e000-6216-11eb-871d-26ad2e7b1e59.png" height="55px">   <img src="https://user-images.githubusercontent.com/25985824/106288518-24bb7680-6216-11eb-8f10-60052c519586.png" height="50px">

</p>

Control FiftyOne datasets through AI assistants using the Model Context Protocol

License PyPI Python

Discord Hugging Face Voxel51 Blog Newsletter LinkedIn Twitter Medium

Documentation · FiftyOne Skills · FiftyOne Plugins · Discord

</div>

What is the FiftyOne MCP Server?

Enable Agents to explore datasets, execute operators, and control the FiftyOne App through natural language. This server exposes 45+ MCP tools across data operations, App UI control, and the full operator/plugin ecosystem.

"List all my datasets"
"Load quickstart dataset and show summary"
"Find similar images in my dataset"

The server starts with 50 built-in operators. Install plugins to expand functionality - the AI can discover and install plugins automatically when needed (brain, zoo, annotation, evaluation, and more).

Available Tools

CategoryToolsDescription
📊 Dataset Management3List, load, and summarize datasets
🎯 App Operations29Control the App UI (views, panels, selection, ...)
⚡ Operator System3Discover and execute any FiftyOne operator
🔄 Pipelines2Run pipelines and manage delegated operations
🔌 Plugin Management5Discover, install, and manage plugins
🖥️ Session1Launch the FiftyOne App server
📈 Aggregations8Count, distinct, bounds, mean, histogram, ...
🧬 Samples5Add, tag, untag, and set values on samples
🗂️ Schema2Inspect and modify dataset field schemas
🎨 App Config6Color scheme, sidebar groups, active fields

Tool modes

45+ tools organized by runtime mode:

  • SDK: Data operations that work everywhere (datasets, aggregations, schema, samples, operators, plugins). No App connection needed.
  • APP: Controls the FiftyOne App UI in real time (set_view, open_panel, notify, select_samples, reload, and 25+ more). Requires a connected browser via ctx.ops.
  • SESSION: Bootstrap tools for starting a local App server (launch_app). Used from terminal environments.

Choosing your tools

Which tools are available depends on how you integrate the server:

IntegrationModesUse case
FiftyOne App pluginapp + sdkAgent panel inside the App (full UI control + data operations)
Terminal / CLIsession + sdkHeadless agent (launch the App, query data, execute operators)

Tool risk levels

Every tool is tagged with a risk level that your agent can use for auto-approval decisions:

  • LOW Safe to auto-execute without prompting (read-only queries, UI state changes)
  • OPERATOR Wraps a FiftyOne operator whose own severity should be checked before executing

Quick Start

Step 1: Install the MCP Server

pip install fiftyone-mcp-server

⚠️ Important: Make sure to use the same Python environment where you installed the MCP server when configuring your AI tool. If you installed it in a virtual environment or conda environment, you must activate that environment or specify the full path to the executable.

Step 2: Configure Your AI Tool

<details> <summary><b>Claude Code</b> (Recommended)</summary>
claude mcp add fiftyone -- fiftyone-mcp
</details> <details> <summary><b>Claude Desktop</b></summary>

Edit ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "fiftyone": {
      "command": "fiftyone-mcp"
    }
  }
}
</details> <details> <summary><b>Cursor</b></summary>

Install in Cursor

Add to ~/.cursor/mcp.json:

{
  "mcpServers": {
    "fiftyone": {
      "command": "fiftyone-mcp"
    }
  }
}
</details> <details> <summary><b>VSCode</b></summary>

Install in VS Code

Add to .vscode/mcp.json:

{
  "servers": {
    "fiftyone": {
      "command": "fiftyone-mcp"
    }
  }
}
</details> <details> <summary><b>ChatGPT Desktop</b></summary>

Edit ~/Library/Application Support/ChatGPT/config.json:

{
  "mcpServers": {
    "fiftyone": {
      "command": "fiftyone-mcp"
    }
  }
}
</details> <details> <summary><b>uvx (No Install Needed)</b></summary>

If you have uv installed:

{
  "mcpServers": {
    "fiftyone": {
      "command": "uvx",
      "args": ["fiftyone-mcp-server"]
    }
  }
}

This downloads and runs the latest version automatically.

</details>

Step 3: Use It

"List all my datasets"
"Load quickstart dataset and show summary"
"Open the map panel and show me the embeddings"
"Select samples with confidence above 0.9"
"What plugins are available? Install the brain plugin"
"Find near-duplicate images in my dataset"

Claude will automatically discover operators and execute the appropriate tools.

Contributing

We welcome contributions! Here's how to set up a local development environment:

  1. Clone the repository

    git clone https://github.com/voxel51/fiftyone-mcp-server.git
    cd fiftyone-mcp-server
    
  2. Install dependencies

    poetry install
    
  3. Run the server locally

    poetry run fiftyone-mcp
    
  4. Test your changes

    poetry run pytest
    poetry run black -l 79 src/
    npx @modelcontextprotocol/inspector poetry run fiftyone-mcp
    
  5. Submit a Pull Request

Resources

ResourceDescription
FiftyOne DocsOfficial documentation
FiftyOne SkillsExpert workflows for AI assistants
FiftyOne PluginsOfficial plugin collection
Model Context ProtocolMCP specification
PyPI PackageMCP server on PyPI
Discord CommunityGet help and share ideas

Community

Join the FiftyOne community to get help, share your ideas, and connect with other users:


<div align="center">

Copyright 2017-2026, Voxel51, Inc. · Apache 2.0 License

</div>

Related MCP servers

Trusted African data, provenance, citations, and catalog search for AI agents.

View repository →
HIHiberden logo

Hiberden

Active

Read and verify your local media archive's 3-2-1 coverage across tape, disk, NAS, and cloud.

0
Dockerfile
MIT
View repository →

Connect Claude, Cursor, or ChatGPT to your business data. Ask questions, get answers.

MCP server for Agon Gateway x402, Helius Wallet, and Tokens API market data.

0
JavaScript
View repository →

Evidence-based register of known Copilot Studio errors and limits, with dated, human-verified fixes.

MCP server for Java call graph generation, static analysis, method chains & config management

4
Java
View repository →