Explorar skills

Canonical web-platform reference — HTML, CSS, JavaScript, DOM/Web APIs, HTTP, SVG, MathML, ARIA, WebAssembly, glossary — fetched from github.com/mdn/content. Auto-activates whenever a specific web feature is named — CSS properties/selectors/at-rules (:has, ::before, @container, color-mix, ...); JS built-ins (Array.flatMap, Promise.allSettled, structuredClone, ...); DOM/Web APIs (fetch, Intersectio

나의 작은 사무실 일꾼들. AI 에이전트 팀이 당신의 직원이 됩니다. 기획 조사, 팀 회의, 의사결정, 보고 영상 제작, 코드 구현까지 — 당신은 사장님처럼 앉아서 결재만 하세요.

Turn a raw Whisper SRT (or any transcript) into clean, well-formatted subtitles and produce a subtitled video. Optimized for spoken-word audio where preserving disfluencies matters.

Provides comprehensive code review guidance for React 19, Vue 3, Rust, TypeScript, Java, Python, and C/C++. Helps catch bugs, improve code quality, and give constructive feedback. Includes a mandatory verification phase: every suspected bug must be cross-referenced against the test suite, callers, and concrete execution traces before being reported. Use when: reviewing pull requests, conducting PR

Generate and repair VS Code-compatible tldraw `.tldr` files from diagram JSON. Use when asked to build or update architecture diagrams from JSON, fix blank tldraw canvas issues in VS Code, or enforce current tldraw schema compatibility.

Deep iterative research. Takes a topic, an optional existing research doc, and a number of iterations. Each iteration spawns a subagent that improves the doc by filling gaps and avoiding duplicate research. The doc has a defined schema (Findings, Investigated Paths, Decision Log, Open Questions) that each subagent knows how to parse and update.

通过 GitHub 仓库 / 用户主页 / Demo 链接 / 简历附件，评估候选人潜力，输出 HR + 技术双视角招聘评估报告。扮演资深技术招聘官，覆盖四个维度：潜力/学习曲线、工程素养/协作、产品思维/商业感、真实性识别（反 AI 一键生成 / 反 fork 改名）。Trigger phrases："评估候选人"、"评估 repo"、"看 GitHub"、"看候选人作品"、"评估 demo"、"招人评估"、"看简历附件"、"/github-as-resume"。

"把脉" — Expert Pulse Check for product decisions. Triggers in two ways: (1) AUTOMATIC — when Claude presents multiple options for a major product decision (product positioning, user experience model, data architecture, pricing strategy, go-to-market, feature prioritization), Claude must automatically enter expert review mode instead of simply asking the user to pick. (2) MANUAL — when the user says

Convene a council of 5 expert agents (PM, QA, Lead Dev, Product Designer, Domain Expert) to deliberate on a topic through anonymous consensus rounds until agreement is reached.

Performs comprehensive FiveM resource security, performance, and compatibility audits. Detects backdoors, RATs, SQL injection, event exploitation, NUI vulnerabilities, supply chain attacks, and malware patterns across ESX, QBCore, QBox, ox_lib, and ND_Core frameworks. Use when asked to: audit FiveM script, review FiveM security, optimize FiveM resource, check FiveM performance, FiveM code review,

Forge new Claude Code skills the right way — proper frontmatter, progressive disclosure via references/, helper scripts for deterministic work, mandatory Gotchas, and the iterate-then-extract meta-process. Use when the user wants to create, write, build, scaffold, or improve a skill.

自动将原型代码转化为带智能标注卡片的静态 HTML 长页面，支持需求自动提炼与零删减还原。