Latest Agent Skills

Batch-process pending LLM Gateway account-contribution requests through the admin API: preflight, validate, issue, patch, and refresh usage with deterministic proxy assignment.

Upgrades the local StaticFlow backend behind a running Pingora gateway on port 39180 using blue/green slots, applying config changes via SIGHUP without restarting the gateway.

Ingests JavaScript-heavy external pages into StaticFlow as standalone local interactive mirrors backed by LanceDB, with bilingual article write-back and localized interactive locales.

Drafts concise, truthful English appeals for GitHub account restrictions, tailored to the user's situation and GitHub Support form fields.

Reposts external blog articles into StaticFlow with style-aware translation, source normalization, image ingestion, and verified publishing.

Imports local Kiro CLI accounts into StaticFlow/llm-access by reading SQLite auth files and calling the local admin API, with proxy assignment and validation.

Resolves an approved article comment task into a publishable AI markdown reply by fetching article context from a local HTTP API (with sf-cli fallback), reasoning with the selected quote and full article content, and writing the final reply to a backend-provided local result file path.

Writes mechanism-first technical implementation deep-dive articles that explain design intent, data flow, trade-offs, and operations, with code locations as supporting evidence.

Ingests music into a LanceDB via ncmdump-rs and sf-cli, searching and downloading from Netease Cloud Music or Bilibili, decrypting local NCM files, and extracting metadata, lyrics, and cover art.

Deploy Caddy on a remote Linux host to expose a local HTTP backend as trusted HTTPS, prioritizing a user-owned domain with Let's Encrypt HTTP-01 and falling back to DuckDNS DNS-01 when needed.

Create high-quality bilingual detailed summaries for StaticFlow articles, focusing on evidence, reader value, and clear structure while allowing flexible writing style.

Publishes Markdown articles and images into LanceDB via the sf-cli tool, syncs notes, infers missing metadata, and verifies results with local API checks.

Drafts GitHub pull request titles, bodies, issue-closing footers, and maintainer ping comments, tailored to repo-specific commitlint or semantic PR conventions.

Rotates GitHub account passwords and logs into security settings using an isolated browser, with manual 2FA handling and optional TOTP automation.

Recalibrates conservative Kiro cache-estimation coefficients from recent successful usage samples, producing a recommendation for manual admin Kmodel updates.

Generates one-day Kiro credit usage reports from LanceDB usage events, with China-time windows, global totals, and account subset breakdowns.

Automates Kiro CLI account onboarding with social Google or GitHub login through an HTTP proxy, including local SQLite auth cleanup, llm-access account import, proxy assignment, balance refresh, and student-tier verification.

Search locally saved Codex session history stored in SQLite and session_index.jsonl, with filters for provider, time range, and archived status.

Automates bulk pickup from plus.keria.cc.cd by submitting codes to a multipart form, downloading the resulting ZIP, and unpacking it locally with a choice of plus_json or sub2api output bundles.

Compacts and prunes LanceDB tables in StaticFlow databases to merge fragment files, reduce open file descriptors, and reclaim storage.

Operates the gpt2api-rs image gateway: start the service, run admin control-plane operations, exercise OpenAI-compatible image APIs, and drive the StaticFlow /admin/gpt2api-rs integration page.

Adds smooth, native-feeling animations to React apps using the View Transition API, covering shared elements, route changes, and list reordering with CSS-based control.

Optimizes React and Next.js performance with 69 prioritized rules covering waterfalls, bundle size, server-side rendering, and re-rendering patterns.

Provides React composition patterns to avoid boolean prop proliferation, covering compound components, state management, and React 19 API changes.

Provides best practices for building performant React Native and Expo apps, covering list performance, animations, navigation, UI patterns, state management, rendering, monorepo structure, and configuration.

Deploy and manage Vercel projects using token-based authentication, avoiding interactive login. Handles deployment, environment variables, and domain management via the Vercel CLI.

Reviews UI code for compliance with Web Interface Guidelines, checking accessibility, design, and UX best practices.

组合搜索、详情、发布、互动等命令，完成小红书竞品分析、热点追踪、内容创作和互动管理等复合运营任务。

通过命令行脚本在小红书上发表评论、回复评论、点赞和收藏笔记，操作前需用户确认内容。

Publishes image-text, video, and long-form content to Xiaohongshu (Little Red Book) via a local CLI script, with support for scheduling, tags, and visibility settings.

小红书内容发现与分析技能。搜索笔记、浏览首页、查看详情、获取用户资料。 当用户要求搜索小红书、查看笔记详情、浏览首页、查看用户主页时触发。

Deploys projects to Vercel as preview deployments, linking to git for automatic deploys, with fallback for unauthenticated environments.

Manages Xiaohongshu (Little Red Book) login state via a CLI script, supporting QR code and phone verification code login, logout, and status checks.

Builds elaborate, multi-component claude.ai HTML artifacts using React, Tailwind CSS, and shadcn/ui, with scripts to initialize and bundle projects into a single HTML file.

Creates animated GIFs optimized for Slack using PIL, with utilities for frame assembly, validation, and easing.

Applies Anthropic's official brand colors and typography to artifacts such as presentations, ensuring a consistent look and feel.

Builds distinctive, production-grade frontend interfaces with bold aesthetic direction, avoiding generic AI design patterns.

Guides users through a structured three-stage workflow for co-authoring documents, from context gathering to reader testing.

Creates algorithmic art in two phases: first writing a generative-art philosophy as a markdown manifesto, then expressing it as interactive p5.js sketches with seeded randomness, flow fields, and particle systems.

Tests local web applications with Playwright, including server lifecycle management, DOM inspection, and screenshot capture.

Writes internal communications such as status reports, leadership updates, 3P updates, company newsletters, FAQs, and incident reports using company-specific formats.

Applies curated color palettes and font pairings to presentation slide decks and other artifacts, with ten pre-set themes and the ability to generate custom themes on demand.

Guides the creation, iterative improvement, and evaluation of agent skills, including drafting, testing, benchmarking, and optimizing descriptions.

Creates original visual art in .png and .pdf documents by first generating a design philosophy and then expressing it visually on a canvas.

Guides the creation of high-quality MCP servers, covering design, implementation, testing, and evaluation for integrating external services with LLMs.

Reference for building Tavily-powered search, extraction, crawling, and research into agentic workflows and RAG systems.

Controls a live browser session on scraped pages to click, fill forms, navigate flows, and extract data using natural language prompts or code, including support for authenticated sessions via profiles.

Saves an entire website as local files by mapping pages and scraping each into organized directories, supporting multiple formats and screenshots.

Searches the web and optionally extracts full page content, returning results as JSON files.

Extracts clean, LLM-optimized markdown from any URL, including JavaScript-rendered SPAs, with support for concurrent scraping of multiple URLs and options like main-content-only extraction and custom output formats.

Bulk extract content from an entire website or site section by crawling pages that follow links, with configurable depth, path filters, and concurrency.

Extracts structured JSON data from complex multi-page websites using an AI agent that navigates pages and returns results matching a schema.

Search the web, code, and company information using the free Exa MCP server, with no API key required.

Automates Xiaohongshu operations including login, content publishing, search, and social interactions via a CLI script.

Translates English text paragraph by paragraph into Chinese and provides deep linguistic analysis covering vocabulary, phrases, grammar, voice, syntax, and sentence structure.

Orchestrates the end-to-end process of turning a resource into learning material by delegating to resource-gatherer, syntax-interpreter, and study-builder.

Discovers and lists all URLs on a website, with optional search filtering to find specific pages within large sites.

Acquires resources from URLs, PDFs, and local files, categorizes them, and organizes content and media into a structured workspace with a lowercase assets folder.

Performs AI-powered web searches with real-time information using Perplexity models via LiteLLM and OpenRouter, providing grounded answers with source citations.

Conduct deep, multi-source web research with citations using the Tavily CLI, producing structured reports for comparisons, market analysis, and literature reviews.