Agile Planner MCP Server (v1.3.3) - AI-Powered Agile Backlog Generator

What is Agile Planner MCP Server?

Agile Planner MCP Server is an MCP-compatible tool that automatically generates complete agile backlogs—including epics, user stories, MVP, and iterations—or specific features from a simple project description. It works directly within Windsurf, Cascade, or Cursor IDEs and requires no technical skills to operate.

How to use Agile Planner MCP Server?

Install the server via npm (npx agile-planner-mcp-server) or point to a local installation, then configure environment variables such as OPENAI_API_KEY or GROQ_API_KEY. Invoke commands like mcp0_generateBacklog or mcp0_generateFeature from within the IDE or use the CLI equivalents (npx agile-planner-mcp-server backlog "Project Name" "description").

Key features of Agile Planner MCP Server

• Generates complete structured agile backlogs from plain English descriptions
• Produces AI-optimized annotations that guide implementation step by step
• Tracks progress with task checkboxes and dependency management
• Organizes output in a dedicated .agile-planner-backlog folder
• Automatically associates features with relevant epics
• Supports both OpenAI and Groq LLMs for backlog generation

Use cases of Agile Planner MCP Server

• Quickly bootstrap an agile backlog for a new SaaS project or mobile app
• Generate a discrete feature with user stories for a specific iteration
• Standardize backlog format across teams without manual formatting effort
• Provide AI coding assistants with precise implementation guidance

FAQ from Agile Planner MCP Server

Can I generate a single feature instead of a full backlog?

Yes. Use the mcp0_generateFeature command and specify the feature description, optional story count, business value, iteration, and epic.

What LLM providers are supported?

The server can use either OpenAI (via OPENAI_API_KEY) or Groq (via GROQ_API_KEY) to generate the backlog structure.

Where are the generated files stored?

All output is placed in a .agile-planner-backlog folder in the current working directory (or a custom path specified via outputPath).

What environment variables are required?

At minimum, set MCP_EXECUTION to "true" and provide either OPENAI_API_KEY or GROQ_API_KEY. Optional variables include DEBUG, TEST_MODE, and AGILE_PLANNER_OUTPUT_ROOT.

Which IDEs or editors are supported?

The server is designed to integrate with Windsurf, Cascade, and Cursor via the MCP protocol, and it can be used from any MCP-compatible client.