Skip to main content
Glama
autoexecbatman

Enhanced Architecture MCP

reasoning_assist

Provides structured reasoning assistance to break down complex problems into logical steps using specified models.

Instructions

Structured reasoning assistance for complex problems

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
problemYesProblem statement requiring reasoning
stepsNoNumber of reasoning steps requested
modelNoModel to use for reasoningarchitecture-reasoning:latest

Implementation Reference

  • The main handler function for 'reasoning_assist' that constructs a structured multi-step reasoning prompt based on the problem and number of steps, then delegates to the local AI via queryLocalAI.
    async reasoningAssist(problem, steps = 5, model = 'architecture-reasoning:latest') { const structuredPrompt = `Problem: ${problem} Please provide a structured reasoning approach with exactly ${steps} steps: 1. [Step 1 reasoning] 2. [Step 2 reasoning] 3. [Step 3 reasoning] ${steps > 3 ? '4. [Step 4 reasoning]' : ''} ${steps > 4 ? '5. [Step 5 reasoning]' : ''} ${steps > 5 ? Array.from({length: steps - 5}, (_, i) => `${i + 6}. [Step ${i + 6} reasoning]`).join('\n') : ''} Conclusion: [Final reasoning conclusion] Think step by step and show your reasoning process clearly.`; return await this.queryLocalAI(structuredPrompt, model, 0.6); }
  • Input schema definition for the reasoning_assist tool, specifying required 'problem' string, optional 'steps' number (default 5), and 'model' string (default 'architecture-reasoning:latest').
    inputSchema: { type: 'object', properties: { problem: { type: 'string', description: 'Problem statement requiring reasoning' }, steps: { type: 'number', description: 'Number of reasoning steps requested', default: 5 }, model: { type: 'string', description: 'Model to use for reasoning', default: 'architecture-reasoning:latest' } }, required: ['problem'] }
  • Tool registration in the ListTools response, defining name, description, and input schema.
    { name: 'reasoning_assist', description: 'Structured reasoning assistance for complex problems', inputSchema: { type: 'object', properties: { problem: { type: 'string', description: 'Problem statement requiring reasoning' }, steps: { type: 'number', description: 'Number of reasoning steps requested', default: 5 }, model: { type: 'string', description: 'Model to use for reasoning', default: 'architecture-reasoning:latest' } }, required: ['problem'] } },
  • Registration of the tool handler in the CallToolRequest switch statement, routing calls to the reasoningAssist method.
    case 'reasoning_assist': return await this.reasoningAssist(args.problem, args.steps, args.model);

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/autoexecbatman/arch-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server