Wassden

An MCP server that converts prompts into a full SDLC workflow: requirements, design, tasks, and code with traceability.
  • typescript

1

GitHub Stars

typescript

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "tokusumi-wassden-mcp": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/tokusumi/wassden-mcp",
        "wassden",
        "start-mcp-server",
        "--transport",
        "stdio"
      ]
    }
  }
}

Wassden is a Spec-Driven Development MCP server that transforms any language model into a systematic development assistant. You can generate requirements, design, tasks, and code with full traceability and validation, while keeping outcomes read-only and auditable.

How to use

You use Wassden by running it through an MCP client that supports MCP workflows. Start the Wassden MCP server in stdio mode to enable prompt-based analysis, design generation, and task creation, then drive the workflow from your client to generate a complete SDLC traceability pipeline.

How to install

Prerequisites you need before starting are Python and a compatible MCP client such as Claude Code or UVX. You will install the Wassden MCP server by pulling its MCP entry point through your client and then starting the server in stdio mode.

# Add Wassden to your MCP client (example with Claude Code)
claude mcp add wassden -- uvx --from git+https://github.com/tokusumi/wassden-mcp wassden start-mcp-server --transport stdio

# Run a sample prompt to test Wassden
uvx --from git+https://github.com/tokusumi/wassden-mcp wassden prompt-requirements --userInput "Create a TODO app"

If you are using a different MCP client such as Cursor, GitHub Copilot, or another UVX-compatible client, configure your client to point to Wassden using the provided MCP entry and execute the start command to run Wassden locally.

Additional configuration and start commands

You can also configure the MCP client to run Wassden via a local setup by editing your MCP configuration file. The following example shows a standard local setup using the same start command as shown for other clients.

{
  "mcpServers": {
    "wassden": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/tokusumi/wassden-mcp",
        "wassden",
        "start-mcp-server",
        "--transport",
        "stdio"
      ]
    }
  }
}

What you can do with Wassden

Wassden enables a complete SDLC workflow: analyze a project description, generate requirements prompts, validate the resulting specs, design architecture, create task breakdowns, and map dependencies. It emphasizes 100% traceability between requirements, design, and tasks.

Security and safety notes

Wassden operates in a read-only mode for the analyzed content, ensuring no file modifications occur during processing. Your prompts and generated outputs are used for analysis and prompt generation only, preserving your source material.

Troubleshooting and tips

If you encounter issues starting Wassden, verify that your MCP client is correctly configured with the Wassden command and to start the server in stdio mode. Ensure you are using a compatible UVX or MCP client and that the repository URL is accessible from your environment.

Notes

For deeper integration guidance, you can explore development-oriented setup and experimental features in development workflows and related tooling. Use the standard start command to ensure compatibility across clients.

Understanding the components and workflow

Core tools drive the Wassden workflow. You generate requirements, then design, then tasks, then code, all with validation and traceability. The system also offers a validation suite to check formats and dependencies and a traceability module to map relationships across requirements, design, and tasks.

Available tools

prompt-requirements

Analyzes input for completeness and generates a requirements prompt. Ensures completeness or prompts for missing details.

prompt-design

Generates architectural design prompts based on the generated requirements to outline system structure.

prompt-tasks

Creates a work breakdown structure prompt from the design to define actionable tasks.

validate-requirements

Validates that generated requirements conform to the EARS format and are complete.

validate-design

Checks architectural consistency and coherence between requirements and design.

validate-tasks

Verifies task dependencies and coverage to ensure a complete implementation plan.

get-traceability

Generates a full dependency map linking requirements, design, and tasks for traceability.

analyze-changes

Performs impact analysis for specification changes across the SDLC artifacts.

generate-review-prompt

Produces task-specific implementation review prompts to guide development.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational