Enhanced Miyabi MCP Server
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Enhanced Miyabi MCP Server generate a medical AI article about LLM diagnostics"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Enhanced Miyabi MCP Server
Manusに与えられたタスクを自動的に振り分け、適切に処理するMCPサーバーです。
機能
タスク意図分析: ユーザーのプロンプトを分析し、開発タスク、記事執筆タスク、一般タスクを判定
自動ディスパッチ: タスクタイプに応じて適切なハンドラーに自動振り分け
開発タスク処理: Miyabi Agentを使用した開発タスクの自動処理
記事生成: 医療AI関連の学術記事を自動生成(LLMO対策、サムネイル指示付き)
Related MCP server: Delia
インストール
cd /home/ubuntu/enhanced-miyabi-mcp
npm installテスト実行
npm test利用可能なツール
1. miyabi__analyze_task_intent
ユーザーのタスクを分析し、タスクタイプを判定します。
入力:
prompt: 分析するユーザーのプロンプト
出力:
isDevelopmentTask: 開発タスクかどうかisArticleWritingTask: 記事執筆タスクかどうかtaskType: タスクタイプ("development", "article", "general")
2. miyabi__auto_dispatch
タスクを自動的に適切なハンドラーに振り分けます。
入力:
prompt: 振り分けるタスクのプロンプト
出力:
action: 実行するアクションmessage: メッセージnextStep: 次のステップ
3. miyabi__handle_development_task
開発タスクを処理します。
入力:
prompt: 開発タスクの内容projectPath: プロジェクトのパス(オプション)
出力:
status: 処理ステータスsteps: 実行されたステップ
4. miyabi__generate_article
医療AI関連の学術記事を生成します。
入力:
articleTitle: 参照元の論文/記事のタイトルarticleAuthor: 著者名articleSource: 出典articleUrl: 参照元のURLkeyConcepts: キーコンセプトperspective: 執筆者のペルソナ・専門分野angle: 論文の切り口deepDivePoints: 特に深掘りしたい論点
出力:
status: 処理ステータスarticle: 記事情報nextSteps: 実行されるステップ
MCP設定
config/mcp-config.jsonを参照してください。
ライセンス
Apache-2.0
Available Tools
4 toolsmiyabi__analyze_task_intentC
ユーザーのタスクを分析し、開発関連か一般タスクかを判定します
Inputs: prompt (required).
miyabi__auto_dispatchB
タスクを自動的に適切なハンドラーに振り分けます(開発タスク→Miyabi、一般タスク→Manus)
Inputs: prompt (required).
miyabi__generate_articleC
医療AI関連の学術記事を生成します(LLMO対策、サムネイル指示付き)
Inputs: articleTitle (required), articleUrl (required), perspective (required), angle (required), +4 more.
miyabi__handle_development_taskC
開発タスクを処理します(Issue作成→Agent実行)
Inputs: prompt (required), projectPath.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 4 tools
The tools have distinct primary purposes: analyzing task intent, dispatching tasks, generating articles, and handling development tasks. However, miyabi__auto_dispatch and miyabi__handle_development_task could be slightly confusing as both involve task handling, but their descriptions clarify dispatch vs. execution roles.
All tools follow a consistent miyabi__verb_noun pattern with underscores, making them predictable and easy to parse. There are no deviations in naming style across the set.
With 4 tools, the count is reasonable but feels borderline thin for a server named 'Enhanced Miyabi MCP Server' that implies broader functionality. The tools cover specific workflows but may leave gaps in a comprehensive task-handling system.
The tools cover key steps like analysis, dispatch, article generation, and development tasks, but there are notable gaps. For example, there's no tool for handling general tasks (Manus) after dispatch, and no update or delete operations for generated articles or tasks, limiting lifecycle coverage.
Maintenance
Related MCP Connectors
Task manager for AI agents. Orchestrate agents to achieve large goals in unified control plane.
AI work orchestration for plans, tasks, teams, and coding-agent dispatch.
Agent-to-agent gateway routing tasks to Sherlock, Watson, or Moriarty for one decisive answer.
Intent execution engine for autonomous agent task routing
Related MCP Servers
- AlicenseCqualityDmaintenanceIntelligent routing layer that analyzes tasks and guides your AI agent to delegate work to specialized tools (Gemini, Aider, Copilot) using rule-based and capability-based routing.3MIT
- FlicenseAqualityNot gradedmaintenanceRoutes AI tasks to appropriate local LLM models (quick, coder, MoE, thinking) with automatic model selection, multi-backend support (Ollama, llama.cpp, Gemini), and parallel processing capabilities.91-
- AlicenseAqualityDmaintenanceIntelligent LLM orchestrator that automatically routes tasks to the most appropriate AI model (Gemini, Qwen, Ollama, LM Studio) based on task characteristics, enabling distributed processing and parallel execution across local and network services.3MIT
- AlicenseNot gradedqualityDmaintenanceAn intelligent task management system that provides structured workflows for AI Agents to plan, decompose, and execute complex programming tasks. It features a dedicated research mode for technical investigations and a task memory function to optimize workflows and avoid redundant coding work.1MIT