sunchendd

19 verified skills38 total stars

vllm-test

Use when testing vLLM performance, running benchmarks, comparing inference configurations, cleaning up GPU environments, or generating performance reports. Activates for benchmarking throughput/latency, configuring vLLM serve parameters, using evalscope or vllm bench, and producing comparison tables.

testing2

find-skills

Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.

data-ai2

humanizer-zh

去除中文文本中的 AI 生成痕迹，使文字更自然。Use when "去AI痕迹", "文本人性化", "humanize", "去除AI写作风格", "让文字更自然". 检测 24 种 AI 写作模式并提供修复建议。

data-ai2

patent-specialist

Expert Patent Specialist Agent for writing high-quality patent disclosures. It analyzes technical documents, identifies core innovations, and generates structured patent documents (Problem, Solution, Effect, Claims) with intelligent innovation extensions (defensive expansion, stability, automation) to increase authorization probability. Use this skill when a user provides technical details and requests a patent draft or disclosure.

tools2

skill-cleanup

清理所有 AI 代理配置目录中的断裂 skill 符号链接。Use when "清理skill", "cleanup skills", "移除无效skill", "删除失效链接", "skill链接清理". 扫描 .agent/.agents/.claude/.opencode/.copilot/.codex/.cursor/.iflow/.openclaw 等目录。

development2

skill-creator

Guide for creating effective AI agent skills. Use when "创建新技能", "write a SKILL.md", "create a skill", "new skill for", "update existing skill". Covers structure, frontmatter, naming conventions, dynamic context injection, and subagent execution.

documentation2

super-wardrobe

每日穿搭建议推送。使用时机：当用户说"今天穿什么"、"帮我搭配衣服"、"穿搭建议"、"wardrobe"。自动获取杭州实时天气，AI 生成上衣/裤子/鞋/帽/手表/配色/是否带伞，邮件+Bark 推送。

data-ai2

vllm-dev

Use when developing vLLM features including speculative decoding (Eagle3, MTP, draft model, suffix, parallel drafting), KV cache optimization (sparsity, offloading, prefix caching), attention backends, and throughput/TPS improvements. Activates for architecture design, implementation, and parameter tuning of inference performance features.

development2

siyuan-note

思源笔记 API 集成与知识库管理。Use when "思源笔记", "SiYuan Note", "知识库管理", "笔记操作", "siyuan API". 支持笔记 CRUD、SQL 查询、批量操作、备份导出。需要 SIYUAN_HOST 和 SIYUAN_TOKEN 环境变量。

development2

patent-search

Patent search via Google Patents (Serper.dev API). Search by keyword, inventor, assignee, or patent number. Use when checking patentability, analyzing competitors' patent portfolios, assessing infringement risks, or researching technology landscapes. Triggered by "专利检索", "patent search", "查专利", "专利布局", or any patent-related query.

development2

github-watcher

GitHub 仓库监控。监控 vllm-ascend 新版本发布、DeepSeek 新仓库创建、以及 vLLM 相关有价值的新 PR，发现更新立即 Bark 推送通知。

data-ai2

vibe-daily

每日 AI 编程工具动态推送。使用时机：当用户说"AI 编程工具动态"、"Claude Code 有什么更新"、"vibe coding 资讯"、"今日 vibe coding"。专注 Cursor/Claude Code/Copilot 更新和技巧，区别于 daily-newsletter（泛科技新闻）。

development2

super-fitness

三个月科学减脂计划每日推送。使用时机：当用户说"今天健身"、"运动计划"、"减脂建议"、"每日健身打卡"、"fitness plan"。177cm/90kg→70kg，每日 AI 生成个性化运动+饮食方案，邮件+Bark 推送。

data-ai2

dev-workflow

通用开发验证工作流，用于 AI/ML 推理服务的开发、测试和验证。支持多种硬件后端（Ascend NPU、GPU）和推理引擎（vLLM、MindIE）。包含完整的需求对齐、代码检查、服务部署、性能测试和结果分析流程。当用户提到"开发"、"测试"、"性能对比"、"服务部署"、"推理验证"等需求时使用此工作流。

development2