Pixeltable MCP Server (Developer Edition)
Pixeltable MCP 服务器(开发者版)
作为 MCP 服务器的多模态 AI 数据基础设施。32 个工具 · 13 个资源 · 6 个提示词,用于表管理、AI/ML 流水线、依赖管理、交互式 REPL 等。
使用 sync 端点 + uvloop 以获得 Pixeltable ≥ 0.5.27 的最佳性能(参见 pyproject.toml)。
快速开始
需要 uv:
curl -LsSf https://astral.sh/uv/install.sh | shClaude Code(最简单) — 只需说:
"Install https://github.com/pixeltable/mcp-server-pixeltable-developer as a uv tool and add it to your MCPs"
手动安装:
uv tool install --from git+https://github.com/pixeltable/mcp-server-pixeltable-developer.git mcp-server-pixeltable-developer
claude mcp add pixeltable mcp-server-pixeltable-developer # Claude Code如果 uv 警告 ~/.local/bin 不在你的 PATH 中,请运行 uv tool update-shell(或将该目录添加到 PATH),以便找到 mcp-server-pixeltable-developer。使用 mcp-server-pixeltable-developer --version 进行检查。
从源码安装:
git clone https://github.com/pixeltable/mcp-server-pixeltable-developer && cd mcp-server-pixeltable-developer
uv sync客户端配置
{
"mcpServers": {
"pixeltable": {
"command": "mcp-server-pixeltable-developer",
"env": {
"PIXELTABLE_HOME": "/Users/{you}/.pixeltable",
"PIXELTABLE_FILE_CACHE_SIZE_G": "10"
}
}
}
}从源码安装 — 使用 "command": "uv" 和 "args": ["run", "--directory", "{repo}", "python", "-m", "mcp_server_pixeltable_stio"]。
用户配置 — ~/.cursor/mcp.json(适用于所有工作区):
{
"mcpServers": {
"pixeltable-developer": {
"command": "mcp-server-pixeltable-developer",
"env": {
"PIXELTABLE_HOME": "/Users/you/.pixeltable"
}
}
}
}如果 Cursor 报告 command not found,请使用 uv tool update-shell / which mcp-server-pixeltable-developer 提供的完整路径,例如 "command": "/Users/you/.local/bin/mcp-server-pixeltable-developer"。
从源码开发此仓库 — 可选的项目 .cursor/mcp.json,以便服务器从你的克隆版本运行(替换路径):
{
"mcpServers": {
"pixeltable-developer": {
"command": "uv",
"args": ["run", "--directory", "/path/to/mcp-server-pixeltable-developer", "python", "-m", "mcp_server_pixeltable_stio"],
"env": {
"PIXELTABLE_HOME": "/Users/you/.pixeltable"
}
}
}
}除非你打算运行两个条目,否则不要定义两次相同的服务器名称(用户 + 项目)。
配置更改后请重启你的客户端。
Related MCP server: dltHub-AI-workbench
测试
使用专用的 PIXELTABLE_HOME 进行测试,以免影响其他目录。
MCP Inspector(交互式)
从克隆版本中,在 uv sync 之后:
export PIXELTABLE_HOME="$HOME/.pixeltable-mcp-test"
uv run mcp dev src/mcp_server_pixeltable_stio/server.py:mcp这将启动服务器并在浏览器中打开 MCP Inspector,这样你无需 IDE 即可调用工具、读取资源并尝试提示词。
快速检查: 工具 pixeltable_check_dependencies,表达式为 openai.chat_completions;工具 execute_python,内容为 print(pxt.__version__);资源 pixeltable://version。
Cursor
在 .cursor/mcp.json 的 env 下添加 PIXELTABLE_HOME(参见上方的客户端配置)。重启 Cursor,确认服务器已连接,然后从 MCP 面板运行一个简单的工具。
CLI 健全性检查(无 JSON-RPC)
mcp-server-pixeltable-developer --version
uv run python list_tools.py--version / --help 会立即退出。list_tools.py 仅打印已注册的工具、资源和提示词(导入检查,非完整的 MCP 会话)。
工具 (32)
类别 | 工具 |
表 |
|
数据 |
|
目录 |
|
配置 |
|
AI/ML |
|
依赖 |
|
类型 |
|
文档 |
|
REPL |
|
日志 |
|
显示 |
|
所有工具均以 pixeltable_ 为前缀(REPL/日志辅助工具除外)。完整的 docstrings 可通过 introspect_function 获取。
资源 (13)
URI | 返回内容 |
| 所有表及其计数 |
| 关于表/视图/快照的信息 |
| 列模式 |
| 所有目录 |
| 目录列表 |
| Pixeltable 版本 |
| 数据存储配置 |
| 可用数据类型 |
| 已注册的 Pixeltable 函数 |
| MCP 工具列表 |
| 工作流指南 |
| 系统和依赖诊断 |
提示词 (6)
pixeltable_usage_guide · getting_started · computer_vision_pipeline · rag_pipeline · video_analysis_pipeline · audio_processing_pipeline
示例
Create a table called movies with title, year, and rating columns → insert sample data → query ratings above 8.5
Add a computed column that runs YOLOX object detection on every image
Check what deps I need for openai.chat_completions(...) → install them
execute_python("print(pxt.list_tables())")文档
pixeltable-skill — 任务路由、API 陷阱(
openai.vision与chat_completions、frame_iterator、similarity(string=...)等)以及与当前 Pixeltable 一致的工作流示例
架构
src/mcp_server_pixeltable_stio/
server.py FastMCP server, tool/resource/prompt registration, uvloop activation
core/
tables.py Table CRUD, views, snapshots, replicas, queries, computed columns
directories.py Directory CRUD, listing, moving
dependencies.py Dependency checking, unified installer, diagnostics
udf.py UDF creation, type system, LLM tool wrappers, MCP connections
helpers.py Config, version, docs search, shared utilities
resources.py Read-only MCP resource handlers
prompt.py Prompt templates for common workflows
repl_functions.py Persistent Python REPL, introspection, package management
canvas_server.py Browser canvas for rich content display故障排除
配置更改后重启你的客户端
需要 Python 3.10+ 和
uvuv tool install后出现command not found: 确保~/.local/bin在PATH中(uv tool update-shell)或通过完整路径调用;使用mcp-server-pixeltable-developer --version进行确认检查
PIXELTABLE_HOME是否指向有效的目录使用
log_bug(...)/generate_bug_report()进行结构化问题跟踪在 github.com/pixeltable/mcp-server-pixeltable-developer 提交问题
This server cannot be deployed
Maintenance
Related MCP Connectors
- OleanderOAuthdev.oleander
The all-in-one data stack for agents. Upload files, run SQL, evolve tables, and render charts.
AI workspace for you, your team, and every agent. Tables, docs (images, 4K video), formulas.
Database for your AI agent. Turn its output into data, docs, skills, and apps you can actually use.
A collaborative substrate over your data: vector, knowledge graph, SQL, geospatial, streaming.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables creating interactive data visualizations from natural language queries using DuckDB for local databases or Databricks for enterprise data warehouses. Supports multiple chart types, CSV imports, SQL queries, and automatic statistical analysis through Claude Desktop.19MIT

dltHub-AI-workbenchofficial
FlicenseAqualityBmaintenanceIt gives Claude Code, Cursor, and Codex structured workflows and an MCP server for building data pipelines with dlt - covering REST API ingestion, transformations, data quality, and exploration.1760-- AlicenseNot gradedqualityFmaintenanceEnables data engineers and BI professionals to perform data pipeline development, quality assurance, visualization, and API integration locally with AI assistance.5MIT
- AlicenseAqualityBmaintenanceLets AI assistants like Cursor and Claude safely explore your database and answer questions about your data in plain English.41MIT