Spark
Sign In
MCP Connector

Enable Structured Reasoning for AI Agents

Adds a structured 'think' tool for step-by-step reasoning, plus optional criticize/plan/search tools in advanced mode.

Connect
Works with tavily
Konstantin Krestnikov
Maintainer?
86
Spark score
out of 100
Updated 4 months ago
Version 1.0.0
Models
universal
Add to Favorites

Why it matters

Enhance AI agent capabilities by integrating structured reasoning and explicit thought logging. This asset allows agents to pause, record thoughts, and utilize advanced tools for complex decision-making and multi-step processes.

Outcomes

What it gets done

01

Implement the 'think' tool for structured reasoning.

02

Provide 'criticize', 'plan', and 'search' tools in advanced mode.

03

Enable tool result analysis and multi-policy environment compliance.

04

Facilitate sequential decision-making in agentic workflows.

Install

Add it to your toolbox

Run in your project directory:

curl -fsSL https://spark.entire.vc/get/vb-think-mcp | bash

Capabilities

Tools your agent gets

think

Adds a thought to the log for structured reasoning without changing the environment or database

criticize

Additional tool available in advanced mode for critical analysis

plan

Additional tool available in advanced mode for planning

search

Additional tool available in advanced mode for searching information

Overview

Think MCP Server

Think MCP Server implements Anthropic's 'think' tool, letting an agent record structured reasoning thoughts, with an advanced mode adding criticize, plan, and search tools. Use it in agentic workflows with tool output analysis, policy compliance, or sequential decision-making. It only logs reasoning, it does not act on the environment.

What it does

Think MCP Server implements the "think" tool described in Anthropic's engineering research on enabling Claude to pause and reason in complex tool-use situations. It lets an agent record an explicit thought during multi-step work without changing the environment or database - the thought is simply appended to the log to support reasoning, backtracking, or policy compliance.

When to use - and when NOT to

Use it in agentic workflows involving tool output analysis, policy-heavy environments where compliance needs verifying, or sequential decision-making where each step builds on the last. In advanced mode, it also adds criticize, plan, and search tools for more structured multi-step reasoning (the search tool requires a Tavily API key). It is a reasoning aid, not a tool that performs actions in the world.

Capabilities

  • think: takes a thought string and appends it to the log for structured reasoning, without altering any external state.
  • Advanced mode adds criticize, plan, and search tools for richer agentic reasoning workflows.

How to install

"mcpServers": {
    "think-mcp": {
        "command": "uvx",
        "args": ["think-mcp"],
        "enabled": true
    }
}

For advanced mode (adds criticize/plan/search):

"mcpServers": {
    "think-mcp": {
        "command": "uvx",
        "args": ["think-mcp", "--advanced"],
        "enabled": true,
        "env": {
            "TAVILY_API_KEY": "your_tavily_api_key_here"
        }
    }
}

Who it's for

Developers building agentic AI workflows who want to improve reasoning quality in complex, multi-step tool-use situations, per Anthropic's research on the think tool.

Source README

Think MCP Tool

Think MCP is an implementation of an MCP (Model Context Protocol) server that provides a "think" tool for structured reasoning in agentic AI workflows. This project is inspired by the Anthropic engineering article: The "think" tool: Enabling Claude to stop and think in complex tool use situations.

According to the referenced article, adding the think tool can lead to improved evaluation metrics by enabling reasoning capabilities even in models that do not natively possess advanced reasoning skills.

Think Tool MCP server

What is the "think" tool?

The "think" tool allows an AI agent to pause and record an explicit thought during complex reasoning or multi-step tool use. It does not change the environment or database, but appends the thought to the log, helping the agent process information, backtrack, or comply with detailed policies.

This approach is especially useful for:

  • Tool output analysis (processing results of previous tool calls)
  • Policy-heavy environments (verifying compliance with guidelines)
  • Sequential decision making (where each step builds on previous ones)

Features

  • Implements the "think" tool as described in Anthropic's research
  • Minimal, standards-based MCP server using mcp[cli]
  • Ready for integration with Claude or other agentic LLMs

Usage

MCP server configuration

Add this MCP server to your facorite agent.

"mcpServers": {
    "think-mcp": {
        "command": "uvx",
        "args": ["think-mcp"],
        "enabled": true
    }
}

Tool definition

The "think" tool is defined as:

  • Input: thought (string) - A thought to think about.
  • Behavior: Appends the thought to the log for structured reasoning.

Advanced mode

Adds aditional tools for your agent:

  • criticize
  • plan
  • search
"mcpServers": {
    "think-mcp": {
        "command": "uvx",
        "args": ["think-mcp", "--advanced"],
        "enabled": true,
        "env": {
            "TAVILY_API_KEY": ... YOUR TAVILY API KEY HERE ...
        }
    }
}

Reference

FAQ

Common questions

Discussion

Questions & comments · 0

Sign In Sign in to leave a comment.