MCP.so
登录

Conductor MCP Server

@jarmstrong158

基本信息

分类

开发工具

许可证

MIT

传输方式

stdio

作者

jarmstrong158

提交者

Jonathan Armstrong

关于 Conductor MCP Server

Local automation platform with a web dashboard for managing scheduled Python/batch tasks. Flask backend, Redis workers, APScheduler, SQLite persistence, live debug log. No terminal required after launch.

Conductor MCP Server 配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "conductor": {
      "command": "python",
      "args": [
        "path/to/Conductor/server.py"
      ]
    }
  }
}

Conductor MCP Server 工具

工具列表暂未提供

当前收录暂时没有可展示的工具列表,不代表服务没有工具。请查阅项目文档了解支持的能力。

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is Conductor?

Conductor is a local task runner that gives Claude direct, persistent control over scheduled workers, multi-step pipelines, and email reports through natural language conversation. It includes a web dashboard at 127.0.0.1:5000 for humans to monitor status, pause tasks, or edit workers manually. Originally built for warehouse reporting automation, it now supports any Python, batch, or shell script with any schedule or chain.

How to use Conductor?

Install via the Windows installer (Conductor_Setup.exe) from the latest release — no Python or Redis required. On first launch, Conductor automatically writes its MCP entry to Claude Desktop’s config; after restarting Claude Desktop once, Claude can control Conductor through conversation (e.g., “Run my metrics script every night at 8:30 PM and email me the report”). The dashboard opens at 127.0.0.1:5000.

Key features of Conductor

  • Claude-controlled automation via 21 MCP tools
  • Flexible scheduling: fixed times, intervals, or cron expressions
  • Supports Python, batch, and shell scripts with visible terminal option
  • Built-in email (Gmail) for reports, alerts, and worker notifications
  • Task chains with parallel execution and stop-on-failure toggle
  • Persistent SQLite state with auto-restart after reboot

Use cases of Conductor

  • Automate daily warehouse reporting by telling Claude the schedule and email recipient
  • Debug a failing uptime check by asking Claude to retrieve recent logs and error details
  • Chain a backup script and a cleanup script to run in sequence every night at 2 AM
  • Use Claude to create a folder watcher that emails files when new ones arrive

常见问题

1

What makes Conductor different from other task schedulers?

Conductor is designed to be controlled entirely through conversation with Claude via the MCP protocol, making automation accessible to non-technical users. It also provides a web dashboard for manual oversight and includes built-in email, templates, and AI error analysis.

2

What dependencies does Conductor require?

On Windows, the installer bundles everything, including Redis. When running from source, Python 3.10+ and a Redis server are required. No other external databases or services are needed.

3

Where does Conductor store its data?

All worker, chain, and group configurations are stored in a local SQLite database (`conductor.db`). Worker state is tracked in Redis, and Gmail credentials are saved in a local `.env` file. Nothing is transmitted externally except to Gmail’s SMTP server.

4

How does Conductor communicate with Claude?

Conductor exposes 21 MCP tools through a bundled MCP server. On first launch, it registers itself with Claude Desktop automatically. No manual API key or transport setup is needed for basic use.

5

Are there any known limits or security considerations?

Conductor runs scripts as subprocesses on the local machine. Selenium or GUI scripts can be run in a visible terminal via `new_console`. Email notifications require a Gmail App Password. The dashboard is accessible only on `127.0.0.1` by default.

评论

开发工具 分类下的更多 MCP 服务器