Frequently Asked Questions (FAQ) about MCP Server for Software Planning

Q: What is an MCP Server, and how does it relate to software planning?

A: MCP stands for Model Context Protocol. An MCP Server acts as a bridge between AI models (like LLMs) and external resources. In software planning, it allows AI to access and interact with planning tools, code repositories, and documentation, enabling smarter and more automated planning processes.

Q: What are the key features of the Software Planning Tool MCP Server?

A: The key features include interactive planning sessions, todo management with complexity scoring, code example integration, and the ability to save and manage detailed implementation plans.

Q: How does complexity scoring help in software planning?

A: Complexity scoring allows you to assign a numerical value (0-10) to each task based on its difficulty. This helps in estimating the time and resources required for each task and overall project, leading to more accurate project timelines.

Q: Can I include code examples in my todo items? Why is this useful?

A: Yes, you can include code examples. This provides immediate context and guidance for developers working on the task, reducing ambiguity and speeding up implementation.

Q: How do I install the Software Planning Tool MCP Server?

A: You can install it either via Smithery (for Claude Desktop) or manually. Smithery offers automated installation, while manual installation provides more control over the configuration. Detailed instructions are available in the documentation.

Q: What tools are available within the MCP Server?

A: The available tools include start_planning, add_todo, get_todos, update_todo_status, save_plan, and remove_todo. Each tool performs a specific function in the software planning process.

Q: How can I use the start_planning tool?

A: The start_planning tool initiates a new planning session and requires a goal parameter, which defines the overall objective of the planning session. For example: { goal: "Create a React-based dashboard application" }

Q: How do I add a new task to my plan?

A: Use the add_todo tool. It requires a title, a detailed description, a complexity score, and an optional codeExample. For example: { title: "Set up project structure", description: "Initialize React project with necessary dependencies", complexity: 3, codeExample: "npx create-react-app dashboard" }

Q: How do I mark a task as complete?

A: Use the update_todo_status tool. It requires the todoId of the task and a boolean value isComplete to indicate whether the task has been finished. For example: { todoId: "some_unique_id", isComplete: true }

Q: How do I save my implementation plan?

A: Use the save_plan tool. It requires the plan text, which contains a structured representation of the project’s strategy and execution guidelines. For example: { plan: "# Dashboard Implementation Plan..." }

Q: Can I remove a task from my plan?

A: Yes, use the remove_todo tool. It requires the todoId of the item to be removed. For example: { todoId: "some_unique_id" }

Q: How does the Software Planning Tool MCP Server integrate with UBOS?

A: The MCP Server is available on the UBOS Asset Marketplace, a curated collection of tools and resources for AI Agent development. UBOS provides a full-stack AI Agent development platform that allows you to orchestrate AI Agents, connect them with your enterprise data, build custom AI Agents, and create multi-agent systems.

Q: Where can I learn more about UBOS and its AI Agent development platform?

A: Visit https://ubos.tech to learn more about UBOS and its capabilities.