M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
S

Shrimp Task Manager

Community
Intelligent task management for AI-powered development with chain-of-thought and reflection.
How to install ↓ Compare GitHub source repository ↗
Category
Dev Tools #355 of 438
Stars
★ 2.1k Popular
Transport
stdio (local process)
Runtime
Node.js 18+
Credentials
No credential needed
License
MIT
Last commit
⚠
Tools
6
40FMRS · D

Shrimp Task Manager is an MCP server for AI agents, emphasizing chain-of-thought and reflection. It provides structured task decomposition, dependency tracking, and persistent memory. It is suitable for integration with MCP-compatible coding assistants, but requires user configuration and setup of a data directory.

Strongest · Setup experience 10/20 Weakest · Security and permissions 7/20

Reliability
8/20
Security and permissions
7/20
Maintenance
8/20
Documentation
7/20
Setup experience
10/20
Why each score
Reliability 8/20
The source shows a server that should start via an entry point (dist/index.js) and implements the MCP protocol using the @modelcontextprotocol/sdk. Tools are registered and handlers are defined clearly. However, there are no committed tests or CI workflows. Thus, the happy path is plausible but not verified automatically, and error handling/edge cases are thin, so capped at 8.
Security and permissions 7/20
No malicious behavior or credential leakage detected. The server stores data in a local directory (DATA_DIR) and may create SQLite databases, but no excessive file access or external network requests were found. However, there is no explicit least-privilege framework or clear confirmation for dangerous operations (e.g., deleting tasks). No red lines are crossed, but security hardening is incomplete, yielding 7.
Maintenance 8/20
The project has an MIT license and is active on GitHub with high stars and relatively recent commits. No releases or changelog are evident. Issue tracking exists, but no security response channel is declared. Thus, maintenance is active but with governance/versioning gaps, giving 8.
Documentation 7/20
The README provides detailed installation and configuration instructions, including Claude Code examples. Many tools and commands are documented, but there is no comprehensive reference for all tool parameters, and no troubleshooting section. Limitations and costs are not clearly disclosed. Thus, usable but with hidden assumptions, scoring 7.
Setup experience 10/20
Clear steps are provided: prerequisites (Node.js), clone, npm install, npm run build, and configure the client. The README includes client configuration examples and hints about permissions. However, manual steps are required and no packaging or one-click install is provided. Nevertheless, steps are clear and easy to follow, with mainstream client examples, so setup scores 10.

Static review · not runListed 2026-08-07

Read the FMRS scoring method →

Fit and risk

What it can accessReads local filesWrites / deletes local filesRuns commands or codeUses the network

Best for

  • AI-assisted development requiring long-term project context.
  • Integration with MCP-compatible coding assistants like Claude Code and Cline.
  • Developers wanting structured task management to increase efficiency.

Not for

  • Simple interactions that don't need task management.
  • Environments with strict security requirements for data location (you must configure DATA_DIR).

Required permissions

  • File system access: needs to read/write task data files (specified by DATA_DIR).
  • Command execution: the MCP server is launched via command line, so the client must allow executing node commands.

Risks and side effects

  • Data persistence: task data is stored locally; regular backup recommended.
  • Misconfiguration can prevent the server from starting.
  • Dependence on the MCP implementation of the AI client may cause compatibility issues.

Setup

Before you start

Runtime：Node.js 18+

DATA_DIR required Path to the directory where task data is stored; required, e.g. /path/to/your/shrimp_data.
Other optional settings (4)
TEMPLATES_USE optional Template language setting, e.g. "en"; optional.
ENABLE_GUI optional Whether to enable the web GUI, "true" or "false"; optional.
WEB_PORT optional Custom port for the web GUI, e.g. 3000; optional.
PROMPT_LANGUAGE optional Prompt language such as en or zh; optional.
  1. Ensure Node.js 18+ and npm/yarn are installed.
  2. Clone the repository: git clone https://github.com/cjo4m06/mcp-shrimp-task-manager.git
  3. cd mcp-shrimp-task-manager
  4. npm install
  5. npm run build
  6. Create a .mcp.json file in your project directory and configure the MCP server (see example in README).
  7. Start your MCP-compatible AI client (e.g., Claude Code) with the config file.
claude_desktop_config.json 
{
  "mcpServers": {
    "shrimp-task-manager": {
      "command": "node",
      "args": [
        "/path/to/mcp-shrimp-task-manager/dist/index.js"
      ],
      "env": {
        "DATA_DIR": "/path/to/your/shrimp_data",
        "TEMPLATES_USE": "en",
        "ENABLE_GUI": "false"
      }
    }
  }
}

Shown for Claude Desktop. Other clients may use a different file or key (VS Code uses "servers") — the configurator below converts it.

.vscode/mcp.json
{
  "servers": {
    "shrimp-task-manager": {
      "command": "node",
      "args": [
        "/path/to/mcp-shrimp-task-manager/dist/index.js"
      ],
      "env": {
        "DATA_DIR": "/path/to/your/shrimp_data",
        "TEMPLATES_USE": "en",
        "ENABLE_GUI": "false"
      }
    }
  }
}

Goes in your project's .vscode/mcp.json (VS Code uses a "servers" key).

Terminal
claude mcp add shrimp-task-manager -e DATA_DIR=/path/to/your/shrimp_data -e TEMPLATES_USE=en -e ENABLE_GUI=false -- node /path/to/mcp-shrimp-task-manager/dist/index.js

Run it in a terminal; replace any <…> placeholders with your own values first.

Generate and validate client config

Check that it works

The shrimp-task-manager tools (init, plan, execute, research, reflect, list) should appear in your client's tool list; typing "plan task: implement user authentication" and getting a generated task plan confirms the connection works.

Troubleshooting

  1. Common checks: 1. Ensure DATA_DIR environment variable is set correctly. 2. Ensure the path to dist/index.js is correct. 3. Ensure Node.js version >=18. 4. Verify the client loads the .mcp.json properly.

Things to try

Once connected, you can ask your AI assistant things like:

  • init project rules
  • plan task: implement user authentication
  • execute task
  • list tasks

Tools 6

init writes
Initialize project rules
plan writes
Create a task plan
execute writes
Execute a specific task
research read-only
Enter research mode
reflect read-only
Review and improve a task
list read-only
List all tasks

Use cases

Feature development: decompose complex features into subtasks and implement step by step.
Bug fixing: analyze issues, create a fix plan, and execute automatically.
Research and learning: systematically compare technologies and create migration plans.

Supported clients

Claude Desktop
Cline

Listed from the project's documentation, not tested by this site.

Overview

Shrimp Task Manager is a task tool built for AI Agents, emphasizing chain-of-thought, reflection, and style consistency. It converts natural language into structured dev tasks with dependency tracking and iterative refinement, enabling agent-like developer behavior in reasoning AI systems. It is an MCP server that provides persistent memory, structured workflows, smart decomposition, and context preservation, helping AI agents avoid losing context or repeating work in software development.

Similar servers

Context7 80 · B

Upstash's official server providing up-to-date third-party library docs for AI coding assistants

★ 63.3k · Tools 2 Compare with this →
Compare with other MCP servers

Source revision 8a3be6bced31 Data synced 2026-10-19 Read the FMRS scoring method