Interact MCP

optimize_user_input

使用配置的 LLM API 来优化或强化用户输入的文本。 此功能可以帮助用户将口语化的、可能存在歧义的输入，转化为更结构化、 更清晰、更便于 AI 模型理解的文本。 Args: original_text: 用户的原始输入文本 mode: 优化模式 - 'optimize': 一键优化，使用预设的通用优化指令 - 'reinforce': 提示词强化，使用用户自定义的强化指令 reinforcement_prompt: 在 'reinforce' 模式下用户的自定义指令 Returns: str: 优化后的文本或错误信息

interactive_feedback

Requests user input via GUI after AI displays complete response in chat. USAGE FLOW: 1. AI displays complete response in chat dialog 2. AI calls this tool to collect user input 3. Tool returns user feedback only This tool collects user input, not for displaying AI responses. AI responses must appear in chat dialog before calling this tool. PARAMETER REQUIREMENTS: - AI MUST provide BOTH 'message' and 'full_response' parameters - Both parameters cannot be empty or whitespace-only - MCP service will automatically select which content to display based on user's display_mode setting USAGE PATTERN: # Step 1: AI displays complete response in chat # Step 2: AI calls tool with BOTH parameters interactive_feedback( message="你希望我实现这些更改吗？", # Required: concise question full_response="我分析了你的代码，发现了3个问题...", # Required: complete response predefined_options=["修复方案A", "修复方案B", "让我想想"] ) Note: MCP service automatically selects appropriate content based on user's display mode configuration.