Ultra MCP

Instructions

Implementation Reference

• src/handlers/ai-tools.ts:401-448 (handler)

  Implementation of the debug-issue tool handler in AIToolHandlers class. Uses ProviderManager to select AI provider, constructs systematic debugging prompt, calls generateText, and returns structured response with metadata.

  async handleDebugIssue(params: z.infer<typeof DebugIssueSchema>) {
    // Use provided provider or get the preferred one (Azure if configured)
    const providerName = params.provider || (await this.providerManager.getPreferredProvider(['openai', 'gemini', 'azure']));
    const provider = await this.providerManager.getProvider(providerName);
    
    const systemPrompt = `You are an expert debugger and problem solver. Help identify and solve technical issues.
    
    Approach debugging systematically:
    - Analyze the problem description and symptoms
    - Identify potential root causes
    - Suggest specific debugging steps
    - Provide solution recommendations
    - Consider edge cases and related issues
    
    Be methodical and provide actionable debugging guidance.`;
  
    let prompt = `Debug the following issue: ${params.task}`;
    if (params.symptoms) {
      prompt += `\n\nSymptoms observed: ${params.symptoms}`;
    }
    if (params.files) {
      prompt += `\n\nRelevant files: ${params.files.join(", ")}`;
    }
  
    const response = await provider.generateText({
      prompt,
      systemPrompt,
      temperature: 0.4, // Balanced temperature for debugging creativity
      reasoningEffort: (providerName === "openai" || providerName === "azure" || providerName === "grok") ? "high" : undefined,
      useSearchGrounding: false, // No search needed for debugging
    });
  
    return {
      content: [
        {
          type: "text",
          text: response.text,
        },
      ],
      metadata: {
        provider: providerName,
        model: response.model,
        symptoms: params.symptoms,
        usage: response.usage,
        ...response.metadata,
      },
    };
  }

• src/server.ts:48-53 (schema)

  Zod schema definition for debug-issue tool input validation, defining task, optional files, symptoms, and provider.

  const DebugIssueSchema = z.object({
    task: z.string().describe("What to debug (e.g., 'fix login error', 'investigate memory leak')"),
    files: z.array(z.string()).optional().describe("Relevant file paths (optional)"),
    symptoms: z.string().optional().describe("Error symptoms or behavior observed"),
    provider: z.enum(["openai", "gemini", "azure", "grok"]).optional().default("gemini").describe("AI provider to use"),
  });

• src/server.ts:304-312 (registration)

  Registration of the debug-issue tool on the MCP server, specifying title, description, input schema, and handler invocation via getHandlers().

  // Register debug-issue tool
  server.registerTool("debug-issue", {
    title: "Debug Issue",
    description: "Debug technical issues with systematic problem-solving approach",
    inputSchema: DebugIssueSchema.shape,
  }, async (args) => {
    const aiHandlers = await getHandlers();
    return await aiHandlers.handleDebugIssue(args);
  });

• src/server.ts:624-641 (registration)

  MCP prompt registration for debug-issue, providing a string-based argsSchema and message construction for prompt-based invocation.

  server.registerPrompt("debug-issue", {
    title: "Debug Issue",
    description: "Debug technical issues with systematic problem-solving approach",
    argsSchema: {
      task: z.string().optional(),
      files: z.string().optional(),
      symptoms: z.string().optional(),
      provider: z.string().optional(),
    },
  }, (args) => ({
    messages: [{
      role: "user",
      content: {
        type: "text",
        text: `Debug this issue: ${args.task || 'Please specify what issue to debug (e.g., error message, unexpected behavior).'}${args.files ? `\n\nRelevant files: ${args.files}` : ''}${args.symptoms ? `\n\nSymptoms observed: ${args.symptoms}` : ''}${args.provider ? ` (using ${args.provider} provider)` : ''}`
      }
    }]
  }));