Skip to main content
Glama
reasoning-tools

reasoning-tools

by MikeyBeez

mcp-reasoning-tools

🧠 Tool-Augmented Reasoning MCP Server for systematic computational verification and enhanced agent capabilities

πŸš€ What This Solves

This MCP server provides computational tools that eliminate calculation errors and format mismatches in agent reasoning. Based on breakthrough research showing 58.3% improvement over baseline reasoning (from 28.6% to 58.3% on BIG-Bench Hard evaluation).

Related MCP server: TinyFn MCP Server

πŸ› οΈ Core Tools

boolean_evaluate

Systematically evaluate boolean expressions with step-by-step verification

Input: "True and False or not True"
Output: Step-by-step boolean evaluation with operator precedence

date_calculate

Perform date arithmetic with computational verification

Input: base_date="2023-01-15", offset_days=7, format="MM/DD/YYYY"
Output: Verified date calculation with breakdown

object_count

Systematically count objects by category with verification

Input: items=["bear", "snake", "microwave", "cat"], target_category="animals"
Output: Categorized count with breakdown (3 animals)

state_track

Track object positions through a series of swaps/moves

Input: initial_state={"Alice": "red"}, operations=[{type: "swap", participants: ["Alice", "Bob"]}]
Output: Step-by-step state tracking

systematic_verify

Apply 6-step systematic reasoning protocol to any problem

Input: problem="Complex reasoning task", problem_type="boolean"
Output: Structured reasoning framework

format_validate

Validate answer format and convert to expected format

Input: answer="True", expected_format="boolean"
Output: Format-validated answer

πŸ“Š Performance Impact

  • Boolean Logic: 100% accuracy with computational verification

  • Date Calculations: 100% accuracy with tool-based arithmetic

  • Object Counting: Systematic categorization prevents errors

  • Format Matching: Eliminates presentation mistakes

  • Overall: 29.7 percentage point improvement on standardized tests

πŸ”§ Installation

  1. Clone and install:

git clone https://github.com/your-username/mcp-reasoning-tools.git
cd mcp-reasoning-tools
npm install
  1. Add to Claude Desktop config:

{
  "mcpServers": {
    "reasoning-tools": {
      "command": "node",
      "args": ["/path/to/mcp-reasoning-tools/src/index.ts"]
    }
  }
}
  1. Restart Claude Desktop

🎯 Usage Examples

Boolean Evaluation

// Instead of mental math: "True and False or not True"
// Use tool: boolean_evaluate({ expression: "True and False or not True" })
// Result: Systematic step-by-step evaluation with verification

Date Arithmetic

// Instead of guessing: "What's 7 days before March 31, 2023?"
// Use tool: date_calculate({ base_date: "2023-03-31", offset_days: -7 })
// Result: Verified calculation with breakdown

Systematic Reasoning

// For any complex problem:
// Use tool: systematic_verify({ problem: "Complex task", problem_type: "temporal" })
// Result: 6-step reasoning protocol applied

πŸ§ͺ Testing

npm test          # Run test suite
npm run lint      # Check code quality
npm run dev       # Start development server

πŸ”¬ Research Foundation

This tool is based on systematic evaluation of agent capabilities using BIG-Bench Hard, a standardized benchmark for challenging reasoning tasks. Key insights:

  • Mental math fails on complex reasoning problems

  • Computational verification eliminates systematic errors

  • Tool augmentation can double agent performance

  • Systematic protocols provide consistent improvements

πŸ“ˆ Integration Strategy

With Existing MCP Tools

  • Complements code execution and web search tools

  • Provides verification layer for computational tasks

  • Systematic reasoning framework for complex problems

With Brain Systems

  • Store reasoning protocols for persistence

  • Track performance improvements over time

  • Learn from successful tool usage patterns

🀝 Contributing

  1. Fork the repository

  2. Create feature branch: git checkout -b feature/enhancement

  3. Add tests for new functionality

  4. Submit pull request with detailed description

πŸ“„ License

MIT License - see LICENSE file for details

πŸ”— Related Projects


🎯 Transform your agent capabilities with systematic computational reasoning!

Available Tools

7 tools
boolean_evaluateA

Systematically evaluate boolean expressions with step-by-step verification

Inputs: expression (required).

date_calculateC

Perform date arithmetic with computational verification

Inputs: base_date (required), offset_days (required), format, operation.

format_validateC

Validate answer format and convert to expected format

Inputs: answer (required), expected_format (required), options.

helpA

Get comprehensive documentation for all reasoning tools functions

Inputs: no named parameters.

object_countC

Systematically count objects by category with verification

Inputs: items (required), target_category (required).

state_trackB

Track object positions through a series of swaps/moves

Inputs: operations (required), initial_state (required).

systematic_verifyC

Apply 6-step systematic reasoning protocol to any problem

Inputs: problem (required), problem_type (required).

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    B3.3/5.0

    Scored across 7 tools

    Disambiguation4/5

    Each tool has a distinct purpose with clear descriptions: boolean_evaluate for boolean expressions, date_calculate for dates, object_count for counting, state_track for tracking moves, systematic_verify for general verification, and format_validate for formatting. The only potential overlap is systematic_verify being general-purpose, but it is clearly distinct from the specialized tools.

    Naming Consistency2/5

    Most tool names follow an object_verb pattern (boolean_evaluate, date_calculate, object_count, state_track, format_validate), but systematic_verify uses adjective_verb and help is a standalone verb. This inconsistency makes the naming convention less predictable.

    Tool Count5/5

    Seven tools is well within the typical 3-15 range for a focused utility server. Each tool covers a distinct reasoning task, and the count feels neither sparse nor bloated.

    Completeness4/5

    The toolset covers several common reasoning tasks (boolean, dates, counting, state tracking, verification, and formatting). The inclusion of systematic_verify and format_validate provides general-purpose coverage that can adapt to other problems, mitigating the need for more specialized tools. Minor gaps exist (e.g., no sequence prediction or arithmetic beyond dates), but they are not critical for the stated purpose.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      D
      maintenance
      Provides a suite of deterministic tools for time calculations, math, and string manipulation that LLMs often struggle to perform accurately. It also includes utilities for secure randomness, data validation, and basic network operations like DNS lookups.
      MIT
    • A
      license
      Not graded
      quality
      C
      maintenance
      Provides over 500 deterministic tools for math, conversions, validation, hashing, and more, enabling AI agents to perform accurate calculations and data transformations without hallucination.
      1
      MIT
    • A
      license
      Not graded
      quality
      D
      maintenance
      Provides deterministic, verifiable text/code/measurement utilities for AI agents, enabling tasks like unit conversion, citation formatting, diffing, proofreading, readability scoring, and syntax checking with re-executable proof.
      14 PyPI
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Provides exact, deterministic tools for math, dates, units, validation, and more to AI agents, returning precise answers with explicit assumptions and warnings instead of model guesses.
      MIT