MCP Server MCP

cascade-img

By github 👁 1 views ▲ 0 votes

LLM-operable Midjourney pipeline: 21 MCP tools to compose, generate, curate with vision, and log.

Homepage Verified source GitHub

Install

uvx cascade-img

Configuration Example

{
  "remotes": [],
  "packages": [
    {
      "registryType": "pypi",
      "registryBaseUrl": "https://pypi.org",
      "identifier": "cascade-img",
      "version": "0.1.1",
      "runtimeHint": "uvx",
      "transport": {
        "type": "stdio"
      },
      "runtimeArguments": [
        {
          "value": "cascade-img",
          "type": "named",
          "name": "--from"
        },
        {
          "value": "cascade-mcp",
          "type": "positional"
        }
      ],
      "environmentVariables": [
        {
          "description": "Discord user-account token the bridge daemon uses to talk to Midjourney (captured from the Discord desktop app; see RUNBOOK.md).",
          "isRequired": true,
          "isSecret": true,
          "name": "DISCORD_USER_TOKEN"
        },
        {
          "description": "Discord channel ID where the Midjourney bot runs.",
          "isRequired": true,
          "name": "MJ_CHANNEL_ID"
        },
        {
          "description": "Version token for Midjourney's /imagine slash command, captured from Discord.",
          "isRequired": true,
          "name": "MJ_IMAGINE_VERSION"
        },
        {
          "description": "URL of the running cascade-mj-bridge daemon that cascade-mcp forwards to.",
          "default": "http://127.0.0.1:5000",
          "name": "CASCADE_BRIDGE_URL"
        }
      ]
    }
  ]
}

mcp model-context-protocol pypi