run-iflow
向助手发出协助请求,适用中华文化理解、中文古文理解、中国特色作品创作、小红书与抖音的宣发文案编写、编写测试用例、信息检索等任务场景,同时也是其他专家助手的强力替补与协助,可以作为大部分认为的备选候补方案。
When & Why to Use This Skill
The run-iflow skill is a versatile MCP-powered orchestration tool that allows Claude to delegate specialized tasks to a variety of external high-performance AI models including GLM, DeepSeek, Qwen, and Kimi. By creating an isolated and clean runtime environment, it enables precise execution of diverse workflows such as Chinese cultural analysis, social media content strategy, and advanced software testing, serving as a robust backup and specialist routing system.
Use Cases
- Chinese Social Media Marketing: Generate high-conversion copy for platforms like Xiaohongshu and Douyin by leveraging GLM-4.6's deep understanding of Chinese market trends and cultural nuances.
- Technical QA and Debugging: Use DeepSeek-V3.2 to automatically generate comprehensive test cases, perform code reviews, and design complex algorithms within an isolated environment.
- Large-scale Information Retrieval: Deploy Kimi-K2-0905 to process massive datasets, summarize long-form documents, and conduct deep-dive knowledge base research.
- Cultural and Academic Research: Interpret ancient Chinese texts and create traditional cultural content with high linguistic accuracy using specialized model presets.
- Workflow Redundancy: Serve as a 'secondary expert' to handle complex requests when primary tools require specialized model capabilities or isolated execution contexts.
| name | run-iflow |
|---|---|
| description | 向助手发出协助请求,适用中华文化理解、中文古文理解、中国特色作品创作、小红书与抖音的宣发文案编写、编写测试用例、信息检索等任务场景,同时也是其他专家助手的强力替补与协助,可以作为大部分认为的备选候补方案。 |
功能简介
通过 MCP 工具向一位外部助手发出协助请求,请对方在一个上下文干净的、运行时隔离的、独立环境中,执行中华文化理解、中文古文理解、中国特色作品创作、小红书与抖音的宣发文案编写、编写测试用例、信息检索等任务。
使用方法
使用 mcp__plugin_headless-knight_runCLI__iflow 工具来向助手发出请求,可传入参数包括:
prompt: string,从上下文整理出的完整的任务描述,必填参数systemPrompt: string,从上下文整理出的需要该助手遵守的系统提示词,用于约束他的行为,可选workDir: string,工作目录,默认为当前目录,可选model: string,指定使用哪个模型,取值为"GLM-4.6"、"Qwen3-Coder-Plus"、"DeepSeek-V3.2"、"MiniMax-M2"或"Kimi-K2-0905"等模型代号,可选,默认为 MiniMax-M2env: object,自定义环境变量,键值对,可选
模型选择
- GLM-4.6: 擅长中文内容创作、中文市场分析、中国文化相关的教育与问答等任务
- Qwen3-Coder-Plus: 擅长阿里云用户或需要进行模型精调和私有化部署等任务
- DeepSeek-V3.2: 擅长与编程相关的任务,如代码生成、Bug 修复、代码解释、算法设计、编写单元测试等
- Kimi-K2-0905: 适用于需要处理大量文本信息的场景,如信息检索、长篇内容总结、知识库问答等
- MiniMax-M2: 适用于上述各项任务
相关环境变量
如果不另外设置,则自动使用当前对话上下文中的配置
IFLOW_API_KEY: Codex CLI 的 API_KeyHTTP_PROXY: HTTP 代理地址HTTPS_PROXY: HTTPS 代理地址ALL_PROXY: 默认代理地址IFLOW_CLI_COMMAND: Codex 命令行地址或名称(确保在 PATH 中包含了路径)NODE_ENV: 环境参数,比如 "development"