1,330 boosters — open source, verified from GitHub, ready to install
1. The daemon auto-starts. It starts on first command. Use to clean up. Use only if you need to pre-warm the browser before issuing commands. 2. Always re-snapshot after page changes. Refs are versioned (). After navigation, form submission, or dynamic content loading, old refs are stale. Run to
1. 分析文稿 - 智能拆解内容，生成页面结构清单 2. 设计风格 - 5种预设风格 + 自定义 + AI智能推荐 3. 批量生成 - 输出符合 PPT 兼容性规范的 SVG 文件
A testing practices guide for Deno projects that helps developers write, organize, and run tests effectively using Deno's built-in testing framework. Useful for teams adopting standardized testing workflows in Claude Code environments.
→ 如果用户已提供，直接确认，不重复询问 → JD 可不完整，有多少给多少；简历没有也能出题（基于 JD 理想候选人画像） 将 JD 内容整理为结构化需求清单（内部使用）：
Безопасно удаляет объект из XML-выгрузки конфигурации. Перед удалением проверяет ссылки на объект в реквизитах, коде и других метаданных. Если ссылки найдены — удаление блокируется. Catalog, Document, Enum, Constant, InformationRegister, AccumulationRegister, AccountingRegister, CalculationRegister,
Graphsignal observes inference workloads from a sidecar process — the profiler. It never shares a process with CUDA: the profiler watches the workload externally via CUPTI, OTLP/gRPC, Prometheus scraping, and NVML. Auto-instrumentation covers vLLM, SGLang, and PyTorch out of the box. Two install pat
当用户对已有暗恋对象 Skill 说以下内容时，进入进化模式： 当用户说 时列出所有已生成的暗恋对象。 本 Skill 运行在 Claude Code 环境，使用以下工具：
Generate images using ModelScope's Tongyi-MAI/Z-Image-Turbo model. Optional: Set to use your own API key.
1. Gather requirements: Ask user for project name, program functionality, required instructions, and frontend features 2. Generate project structure: Create complete directory tree following Anchor conventions 3. Create Rust smart contract: Generate lib.rs, state.rs, errors.rs, and instruction handl
This skill gives you full access to a running marimo notebook. You can read cell code, create and edit cells, install packages, run cells, and inspect the reactive graph — all programmatically. The user sees results live in their
为 Pi/OpenClaw agent 提供双引擎网络搜索能力：Grok AI 搜索（实时联网 + AI 分析）+ Tavily 搜索（结构化结果 + 网页抓取）。 在 Bash 中调用以下脚本（确保已加入 PATH 且已 source .env）： 对于 Level 2+ 的复杂搜索，在执行前进行结构化规划：
List installed skills and their status, or create new skills. On invocation, scan skills and present: When user selects a skill:
每次调用本 skill 时，在开始正式任务前优先执行： 该脚本负责快速检查并同步当前 skill 项目；如果更新失败、超时、无网络、无法获取远端状态，或 目录不是 Git 仓库，直接跳过更新并继续执行用户任务，不要卡住。 如果 有文件冲突，脚本会自动将远端最新文件覆盖到本地冲突文件，但不会执行 之类的全局破坏性操作。
Skyll enables agents to dynamically discover and retrieve skills at runtime. Instead of having all skills pre-loaded, agents can search for relevant skills on-demand and inject them into context. Get a specific skill: The API returns ranked skills with relevance scores (0-100):
Automates conversion of product requirement documents (PRDs) into Ralph's prd.json format for autonomous agent execution. Developers managing Ralph-based autonomous systems benefit from streamlined PRD structuring and standardization.
底层原则：先定角色，再拆内容，最后生成图解。 本 Skill 不是头像生成器，也不是自由风格生图器。它的目标是把用户授权提供的个人材料转成一套可长期复用的极简手绘图解角色，再用这个角色和执行 Agent 一起参与内容表达。 推荐在支持图片读取、较高推理强度和图像生成工具的 Agent 环境中使用。这样可以直接完成图片读取、角色建档、三张角色资产生成、内容图解生成和返修。
首次激活时，必须说一次免责声明：「我以Naval视角和你聊，基于公开言论推断，非本人观点」。此后对话绝不重复——重复 = 破坏沉浸感 = 失败。 用户说「退出 / 切回正常 / 跳出角色 / 不用扮演了 / 别演了」中任一关键词 → 立即恢复正常助手语气，不再用「我」自称 Naval，不再用 Oracle 模式短句格言，回到标准助手语气。 1. 这个机会用的是哪种杠杆：劳动/资本/代码/媒体？（搜索商业模式、产品形态）
Quick Codebase Analysis is a fast alias for Gemini codebase analysis that lets developers analyze directories with optional scope filters (e.g., `/c ./src security`). It's useful for developers who need rapid codebase insights without full context overhead.
Mission: own the target, pixel for pixel. Read every value off the live page (never a screenshot guess), rebuild it, prove the match with measured computed-style assertions, then make it yours. This skill runs as a workflow inside a live Claude session, not a background script. When it's invoked, YO
Structured approach to improving performance through focused effort, feedback, and continuous refinement. Based on psychologist K. Anders Ericsson's research on expertise acquisition.
The 'verify' booster helps developers quickly diagnose and fix build failures, test errors, lint issues, and CI pipeline breaks. It's useful for teams using Claude Code who need rapid error recovery and review fix application without bogging down implementation work.
A systematic debugging skill for web applications that leverages Chrome DevTools and Playwright MCPs to troubleshoot APIs, authentication flows, and UI issues. Developers working with web apps will benefit from its structured debugging approach and intelligent context discovery.
This skill must use the Image 2 model to generate every final slide image, thumbnail board, visual page, infographic, article illustration, or PPT page image. HTML output is strictly forbidden for image generation. Do not create HTML/CSS pages, browser screenshots, SVG/HTML mockups, canvas renders,
Guides GitHub for parasite SEO, GEO (AI citation), and curated list creation. GitHub is a Tier 2 Technical Authority platform—high domain authority, fast indexing, very high AI citation probability. Use for repos, README, GitHub Pages, gists, and Awesome-style navigation lists.
