@tosin2013/mcp-shrimp-task-manager/dist/prompts/generators/queryTask.js

Enhanced MCP Shrimp Task Manager with comprehensive LLM integration. A task management tool built for AI Agents, emphasizing chain-of-thought, reflection, and style consistency. Features real GPT-4 ↔ MCP tools communication, comprehensive testing pipeline

/**
 * queryTask prompt 生成器
 * 負責將模板和參數組合成最終的 prompt
 */
import { loadPrompt, generatePrompt, loadPromptFromTemplate, } from "../loader.js";
/**
 * 獲取 queryTask 的完整 prompt
 * @param params prompt 參數
 * @returns 生成的 prompt
 */
export function getQueryTaskPrompt(params) {
    const { query, isId, tasks, totalTasks, page, pageSize, totalPages } = params;
    if (tasks.length === 0) {
        const notFoundTemplate = loadPromptFromTemplate("queryTask/notFound.md");
        return generatePrompt(notFoundTemplate, {
            query,
        });
    }
    const taskDetailsTemplate = loadPromptFromTemplate("queryTask/taskDetails.md");
    let tasksContent = "";
    for (const task of tasks) {
        tasksContent += generatePrompt(taskDetailsTemplate, {
            taskId: task.id,
            taskName: task.name,
            taskStatus: task.status,
            taskDescription: task.description.length > 100
                ? `${task.description.substring(0, 100)}...`
                : task.description,
            createdAt: new Date(task.createdAt).toLocaleString(),
        });
    }
    const indexTemplate = loadPromptFromTemplate("queryTask/index.md");
    const prompt = generatePrompt(indexTemplate, {
        tasksContent,
        page,
        totalPages,
        pageSize,
        totalTasks,
        query,
    });
    // 載入可能的自定義 prompt
    return loadPrompt(prompt, "QUERY_TASK");
}
//# sourceMappingURL=queryTask.js.map