Shrimp Task Manager is a MCP server that turns natural language into structured tasks with dependency tracking and memory across sessions for AI-powered development.
Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).
What it is
Shrimp Task Manager is an MCP (Model Context Protocol) server that organizes AI-assisted software development by preserving context, decomposing tasks, and tracking dependencies.
How it works
The project provides a server that facilitates persistent memory, structured workflows, smart task decomposition, and context preservation across sessions. It includes a web-based task viewer and an optional lightweight web GUI, with a configuration model using .mcp.json to integrate with Claude Code or other MCP clients.
Getting started
Prerequisites include Node.js 18+ and an MCP-compatible AI client. Installation steps cover cloning the repository, installing dependencies, and building:
# Clone the repository
git clone https://github.com/cjo4m06/mcp-shrimp-task-manager.git
cd mcp-shrimp-task-manager
# Install dependencies
npm install
# Build the project
npm run build
Configure Claude Code via a .mcp.json file in your project, and start Claude Code with the specified command:
claude --dangerously-skip-permissions --mcp-config .mcp.json
Recent releases
No releases are listed in the provided data.
Traction
Stars: 2142 Forks: 252 Open issues: 41
License
MIT
What it does
- Keeps tasks and progress across sessions
- Automatically breaks down complex tasks into subtasks
- Tracks dependencies between tasks
- Preserves context across token limits
Configuration
Environment and command usage examples are provided, including how to set DATA_DIR, ENABLE_GUI, and other MCP server settings, as well as available commands like init project rules, plan task, execute task, and continuous mode.
Note: All quoted commands and JSON snippets are taken directly from the README sections above.