agenthold

Name: agenthold
Author: edobusy

edobusy

Shared versioned state for multi-agent AI workflows. An MCP server.

Setup

Choose your client, copy the config, then fill in any required environment variables.

claude mcp add agenthold -- uvx agenthold

~/.codex/config.toml

[mcp_servers.agenthold]
command = "uvx"
args = ["agenthold"]

# codex mcp add agenthold -- uvx agenthold

~/.cursor/mcp.json

{
  "mcpServers": {
    "agenthold": {
      "command": "uvx",
      "args": [
        "agenthold"
      ]
    }
  }
}

.vscode/mcp.json

{
  "servers": {
    "agenthold": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "agenthold"
      ]
    }
  }
}

~/.gemini/settings.json

{
  "mcpServers": {
    "agenthold": {
      "command": "uvx",
      "args": [
        "agenthold"
      ]
    }
  }
}

claude_desktop_config.json

{
  "mcpServers": {
    "agenthold": {
      "command": "uvx",
      "args": [
        "agenthold"
      ]
    }
  }
}

.trae/mcp.json

{
  "mcpServers": {
    "agenthold": {
      "command": "uvx",
      "args": [
        "agenthold"
      ]
    }
  }
}