Open MCP Server

Implementation Reference

• src/tools/summarize-tool.ts:21-34 (handler)

  The handler function for the 'summarize' tool. It extracts parameters, calls generateSummary, wraps in error handling, and returns MCP-formatted text content.
  async ({ text, maxLength, language }) => { return wrapToolExecution(async () => { const options: SummarizeOptions = { text, maxLength, language }; const summary = await generateSummary(options); return { content: [{ type: "text" as const, text: summary }] }; }, { errorCode: ERROR_CODES.AI_OPERATION, context: "Failed to generate summary", rethrow: true }); }
• src/tools/summarize-tool.ts:16-20 (schema)

  Zod schema for input validation of the summarize tool: text (required string), maxLength (optional number), language (optional string).
  { text: z.string().min(1), maxLength: z.number().optional().default(DEFAULTS.MAX_SUMMARY_LENGTH), language: z.string().optional().default(DEFAULTS.SUMMARY_LANGUAGE) },
• src/tools/summarize-tool.ts:14-36 (registration)

  The registration function that defines and registers the 'summarize' tool with the MCP server, including schema and handler.
  export function registerSummarizeTool(server: McpServer): void { server.tool("summarize", { text: z.string().min(1), maxLength: z.number().optional().default(DEFAULTS.MAX_SUMMARY_LENGTH), language: z.string().optional().default(DEFAULTS.SUMMARY_LANGUAGE) }, async ({ text, maxLength, language }) => { return wrapToolExecution(async () => { const options: SummarizeOptions = { text, maxLength, language }; const summary = await generateSummary(options); return { content: [{ type: "text" as const, text: summary }] }; }, { errorCode: ERROR_CODES.AI_OPERATION, context: "Failed to generate summary", rethrow: true }); } ); }
• src/tools/summarize-tool.ts:38-52 (helper)

  Helper function that generates the summary using Google Gemini 1.5 Pro AI model via the 'ai' SDK.
  async function generateSummary(options: SummarizeOptions): Promise<string> { const { text, maxLength, language } = options; const prompt = `Please summarize the following text in ${language}, keeping the summary within ${maxLength} characters:\n\n${text}`; const model = google.chat("gemini-1.5-pro"); const result = await generateText({ model: model, prompt: prompt, maxTokens: maxLength, temperature: DEFAULTS.AI_TEMPERATURE }); return result.text; }
• src/index.ts:68-68 (registration)

  Invocation of the registerSummarizeTool during server initialization to register the tool.
  registerSummarizeTool(server);
• src/types/common.ts:169-173 (schema)

  TypeScript interface defining the shape of SummarizeOptions used in the tool implementation.
  export interface SummarizeOptions { text: string; maxLength: number; language: string; }