aradotso/trending-skills
153 skills165k installs10.1k starsGitHub
Install
npx skills add https://github.com/aradotso/trending-skillsSkills in this repo
1Openclaw Control CenterThe openclaw-control-center skill deploys a local-first security-first dashboard for OpenClaw agents at http://127.0.0.1:4310. Sections include overview system health, usage token spend and context pressure, staff active execution status, collaboration parent-child handoffs, tasks board with gated approvals, memory per-agent health and editing, documents from source files, and settings wiring status. Installation clones the repo, runs npm install, copies .env.example, builds, tests, and starts npm run dev:ui. Security defaults keep READONLY_MODE true, LOCAL_TOKEN_AUTH_REQUIRED true, IMPORT_MUTATION_ENABLED false, and APPROVAL_ACTIONS_ENABLED false with dry-run defaults. TypeScript connectors cover Monitor, StaffConnector, CollaborationTracer, UsageConnector, MemoryConnector, WiringChecker, and gated TaskConnector approve endpoints. Critical constraint: modify only control-center/ directory, never ~/.openclaw/openclaw.json directly. UI navigation uses ?section=overview|usage|staff|collaboration|tasks|memory|documents|settings query params with lang zh or en.4.3kinstalls2Inkos Multi Agent Novel Writinginkos-multi-agent-novel-writing documents InkOS, a multi-agent CLI that autonomously writes, audits, and revises long-form fiction with human review gates. Install via npm global @actalk/inkos or npx, requiring Node.js 20 or newer. Projects store outlines, book rules, chapter summaries, subplot boards, emotional arcs, character matrices, and per-chapter markdown under a story directory created by inkos book create. The pipeline runs Architect, Writer, Validator, Auditor, and Reviser agents, with Validator applying eleven deterministic rules at zero LLM cost before LLM audit across thirty-three dimensions. Commands cover inkos write next, audit, revise in spot-fix rewrite or polish modes, genre management, style import, spinoff canon import, AIGC detection, and a daemon for continuous production. Configuration lives in ~/.inkos/config.json for LLM provider, daemon intervals, webhooks, and optional GPTZero detection. TypeScript examples show programmatic BookManager and agent usage plus Express webhook verification for chapter-complete events.2.7kinstalls3Claude Hud StatuslineClaude HUD is a Claude Code plugin that adds a persistent terminal statusline showing real-time context window usage, active tool calls, running subagents, todo progress, subscriber usage limits, and git branch status. Requirements are Claude Code v1.0.80+ and Node.js 18+ or Bun. Installation adds the jarrodwatts/claude-hud marketplace, installs the claude-hud plugin, runs /claude-hud:setup to write statusLine config, then restarts Claude Code. The default layout shows model name, project path, context bar, and usage bar. Optional expanded lines display live tool edits, agent activity, and todo completion counts. Configuration via /claude-hud:configure offers Full, Essential, and Minimal presets, or manual edits to config.json controlling lineLayout, pathLevels, elementOrder, gitStatus options, display toggles, and colors. Claude HUD uses the native statusline API feeding stdin JSON to claude-hud and parsing transcript JSONL for tool and agent activity, updating about every 300ms with token data from Claude Code rather than estimates.2.6kinstalls4Understand Anything Knowledge GraphThe understand-anything-knowledge-graph skill installs and runs the Understand Anything Claude Code plugin that maps files, functions, classes, and dependencies into knowledge-graph.json with plain-English summaries. Commands include /understand for the five-agent pipeline, /understand-dashboard for a React Flow browser UI, /understand-chat for NL queries, /understand-diff for uncommitted change impact, /understand-explain for file deep dives, and /understand-onboard for team guides. The pipeline runs project-scanner, up to three concurrent file-analyzers, architecture-analyzer layer grouping, tour-generator, and graph-reviewer integrity checks. Output persists under .understand-anything/knowledge-graph.json with incremental re-runs on changed files only. The dashboard uses React Flow, Zustand, Fuse.js search, and persona modes for junior dev, PM, or power user detail levels. Core package APIs support loadKnowledgeGraph, searchGraph, and buildTour programmatically. Install via Claude plugin marketplace or pnpm monorepo build from source. Troubleshoot timeouts on very large repos, stale graphs after refactors by deleting the cache, and dashboard build prerequisites.2.3kinstalls5Agency Agents Ai SpecialistsAgency Agents AI Specialist Personalities Skill by ara so https ara so Daily 2026 Skills collection A curated collection of 50 specialized AI agent personalities for Claude Code Cursor Aider Windsurf Copilot and more Each agent has deep domain expertise a distinct personality defined workflows and measurable deliverables covering engineering design marketing sales paid media and beyond bash git clone https github com msitarzewski agency agents git cd agency agents bash Copy all agents to Claude s agents directory cp r agency agents claude agents The agency agents ai specialists agent skill provides documented workflows prerequisites triggers and safety guidance from its SKILL md source Agents load it when user requests match the description and follow step by step instructions without inventing capabilities It integrates with standard agent tooling for the tasks inputs outputs and failure modes described in the repository documentation2kinstalls6Agent Browser AutomationSkill by ara so https ara so Daily 2026 Skills collection agent browser is a headless browser automation CLI built in Rust designed for AI agents It wraps Chrome via the Chrome DevTools Protocol CDP and exposes a fast ergonomic command line interface for navigation interaction accessibility snapshots screenshots network interception and more with no Node js or Playwright runtime required Recommended npm global bash npm install g agent browser agent browser install Download Chrome for Testing first time only macOS Homebrew bash brew install agent browser agent browser install Rust Cargo bash cargo install agent browser agent browser install The agent browser automation agent skill provides documented workflows prerequisites triggers and safety guidance from its SKILL md source Agents load it when user requests match the description and follow step by step instructions without inventing capabilities It integrates with standard agent tooling for the tasks inputs outputs and failure modes described in the repository documentation1.9kinstalls7Daily Stock AnalysisSkill by ara so https ara so Daily 2026 Skills collection LLM powered stock analysis system for A share Hong Kong and US markets Automatically fetches quotes news and fundamentals generates AI decision dashboards with buy sell targets and pushes results to WeChat Feishu Telegram Discord Email on a schedule via GitHub Actions zero server cost AI Decision Dashboard One line conclusion precise buy sell stop loss prices checklist per stock Multi market A shares CN HK stocks US stocks indices SPX DJI IXIC Data sources AkShare Tushare YFinance for quotes Tavily SerpAPI Brave for news LLM backends Gemini OpenAI Claude DeepSeek Qwen via LiteLLM unified Push channels WeChat Work Feishu Telegram Discord DingTalk Email PushPlus Automation GitHub Actions cron schedule no server needed Web UI Portfolio management history backtesting Agent Q A Agent Multi turn strategy Q A with 11 built in strategies MA crossover Elliott Wave etc Method 1 GitHub Actions Recommended Zero Cost Step 1 Fork the repository https github com ZhuLinsen daily_stock_analysis1.9kinstalls8Ui Ux Pro Max SkillSkill by ara so https ara so Daily 2026 Skills collection UI UX Pro Max is an AI skill that injects design intelligence into coding agents giving them 161 industry specific reasoning rules 67 UI styles 57 font pairings 161 color palettes and pre delivery checklists to produce professional accessible conversion optimized interfaces on the first attempt bash Install the CLI globally npm install g uipro cli Add the skill to your project npx uipro cli install Or install globally npx uipro cli install global The ui ux pro max skill agent skill provides documented workflows prerequisites triggers and safety guidance from its SKILL md source Agents load it when user requests match the description and follow step by step instructions without inventing capabilities It integrates with standard agent tooling for the tasks inputs outputs and failure modes described in the repository documentation1.9kinstalls9Everything Claude Code HarnessThe everything-claude-code-harness skill installs a production agent harness performance system evolved from daily real-world use across Claude Code, Cursor, Codex, OpenCode, and Antigravity. It provides specialized subagents, reusable skills, slash commands, memory-persisting hooks, security scanning, and language-specific rules via plugin marketplace or manual clone install. Directory structure spans agents, commands, skills, hooks, rules per language, and install.sh for targeting TypeScript, Python, Golang, and Swift stacks. Key commands include plan, tdd, e2e, code-review, security-scan, and build-fix in namespaced or short forms. Rules must be installed manually into .claude/rules because plugins cannot auto-distribute them. Hooks cover SessionStart, Stop, PostEdit, and related lifecycle events. Use when setting up ECC plugin, optimizing Claude Code setup, adding skills commands hooks memory, or improving agent harness performance with custom rules and security scanning. The workflow follows the source SKILL.md contract with progressive reference loading, clear trigger phrases, and practical steps developers can apply directly in agent sessions.1.8kinstalls10Shannon Ai PentesterShannon is an autonomous white-box pentester for web apps and APIs. It reconnoiters targets with Nmap, Subfinder, WhatWeb, and Schemathesis, reads repository code to map attack surfaces, then runs parallel exploit agents for SQLi, XSS, SSRF, auth bypass, and authorization flaws. Only confirmed reproducible findings with copy-paste proof-of-concepts reach the final report. Setup requires Docker and Anthropic, Claude Code OAuth, Bedrock, or Vertex credentials. CLI covers start with URL and REPO paths, logs, status, resume via WORKSPACE, stop, and report. Configuration supports rate limits, concurrency, and model selection. Use when teams need automated security testing tied to real source context rather than black-box scans alone.1.6kinstalls11Gstack Workflow AssistantThe gstack-workflow-assistant skill helps with Team of specialist AI workflows for Claude Code with CEO review, engineering planning, code review, shipping, QA testing, and browser automation. Key workflows include gstack Workflow Assistant; What It Does; Installation; Requirements. Documented capabilities cover **CEO Review**: Rethink problems and find the 10-star product hiding in requests; **Engineering Planning**: Lock in architecture, data flow, and edge cases; **Code Review**: Paranoid staff engineer-level review that catches production bugs; **Release Management**: One-command shipping with tests and PR creation; **QA Testing**: Automated browser testing with screenshots and systematic coverage. Agents should invoke it when users ask about gstack workflow assistant or mention triggers defined in the skill frontmatter. Follow the SKILL.md steps, reference files, and output formats rather than improvising outside documented scope. ## What It Does gstack provides specialized AI personas and workflows: - **CEO Review**: Rethink problems and find the 10-star product hiding in requests - **Engineering Planning**: Lock in architecture, data flow, and edge cases - **Code Review**:.1.6kinstalls12Deepseek OcrThe deepseek-ocr skill documents DeepSeek-OCR vision-language OCR with Contexts Optical Compression supporting native and dynamic resolutions. Setup uses conda Python 3.12.9 PyTorch 2.6.0 CUDA 11.8 cloning deepseek-ai DeepSeek-OCR. Inference runs via vLLM throughput or HuggingFace Transformers. Prompt modes include document-to-markdown free OCR figure parsing and grounding outputting structured text or markdown from images and PDFs. Developers invoke when extracting text from documents converting PDFs to markdown or running high-throughput OCR pipelines requiring GPU CUDA stack on local or server inference hosts. DeepSeek-OCR vision-language optical compression. vLLM or HuggingFace inference backends. Modes: doc-to-markdown free OCR figure parsing. Processes images and PDFs to markdown. CUDA 11.8 PyTorch 2.6 Python 3.12.9 setup. DeepSeek-OCR setup and inference. User asks deepseek ocr or document markdown.1.6kinstalls13Autoresearchclaw Autonomous ResearchAutoResearchClaw is a fully autonomous 23-stage pipeline turning natural language topics into academic papers with real arXiv and Semantic Scholar citations, sandboxed experiments, statistical analysis, multi-agent peer review, and conference-ready LaTeX for NeurIPS, ICML, or ICLR. No hallucinated references or human babysitting per skill claims. Installation and setup steps configure the pipeline stages from ideation through experiments, review, and export. Use when users want autonomous paper generation, research pipeline setup, or chat-an-idea-get-a-paper workflows with verifiable citations. 23-stage autonomous pipeline from topic to conference-ready LaTeX paper Real arXiv and Semantic Scholar citations without hallucinated references Sandboxed experiments, statistical analysis, and multi-agent peer review NeurIPS, ICML, ICLR LaTeX output formats supported Triggers on autonomous research, paper generation, and pipeline setup requests AutoResearchClaw runs a 23-stage autonomous research pipeline producing cited academic papers in LaTeX Conference-ready LaTeX paper with verified citations, experiments, and peer review stages User mentions AutoResearchClaw, autonomous research, or.1.5kinstalls14Chrome Cdp Live BrowserThe chrome cdp live browser skill Give AI agents access to your live Chrome session via CDP - interact with open tabs, logged-in accounts, and current page state. Documentation covers workflows, commands, and guardrails agents should follow when users invoke this capability. Key documented areas include "control my chrome browser"; "interact with open chrome tabs"; "automate my live browser session"; "take a screenshot of a webpage". Reference commands include pi install git:github.com/pasky/chrome-cdp-skill@v1.0.1; git clone https://github.com/pasky/chrome-cdp-skill. Use when developers or agents need structured guidance for chrome cdp live browser tasks with evidence grounded in the bundled SKILL.md rather than generic advice. "control my chrome browser" "interact with open chrome tabs" "automate my live browser session" "take a screenshot of a webpage" "click elements on a webpage" "read content from a browser tab" "navigate to a URL in chrome" "use chrome devtools protocol" Give AI agents access to your live Chrome session via CDP — interact with open tabs, logged-in accounts, and current page state1.5kinstalls15Lightpanda BrowserThe lightpanda browser skill Expert skill for Lightpanda - the headless browser built in Zig for AI agents and automation. 9x less memory, 11x faster than Chrome. Installation, CLI, CDP server, Playwright/Puppeteer integration, and web scraping. Documentation covers workflows, commands, and guardrails agents should follow when users invoke this capability. Key documented areas include lightpanda; headless browser for AI; lightweight headless browser; CDP automation. Reference commands include chmod a+x ./lightpanda; chmod a+x ./lightpanda. Use when developers or agents need structured guidance for lightpanda browser tasks with evidence grounded in the bundled SKILL.md rather than generic advice. lightpanda headless browser for AI lightweight headless browser CDP automation web scraping automation zig browser lightpanda browser fast headless browser Expert skill for Lightpanda — the headless browser built in Zig for AI agents and automation. 9x less memory, 11x faster than Chrome. Installation, CLI, CDP server, Playwright/Puppeteer integration, a1.5kinstalls16Gsd 2 Agent FrameworkGSD 2 is a CLI-driven agent framework that transforms a project roadmap and spec into running software by orchestrating Claude through a hierarchy of milestones, slices, and tasks. Each task dispatches to a fresh context window with pre-inlined project context, research findings, and architectural decisions. The system manages git worktree isolation per milestone, crash recovery via lock files, budget controls, stuck detection (retry with diagnostics), and cost ledger tracking. Developers define work in ROADMAP.md and PROJECT.md, then run /gsd auto to walk away while the agent researches, plans, and executes code with verifiable must-haves at each task boundary.1.5kinstalls17Marketingskills Ai Agentsmarketingskills-ai-agents is an agent skill from aradotso/trending-skills that install and use marketing skills for ai agents — cro, copywriting, seo, analytics, and growth engineering skills for claude code and other coding agents. # Marketing Skills for AI Agents > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. [coreyhaines31/marketingskills](https://github.com/coreyhaines31/marketingskills) is a collection of markdown-based AI agent skills covering CRO, copywriting, SEO, analytics, paid ads, email, and growth engineering. Install them once and any compa Developers invoke marketingskills-ai-agents during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.4kinstalls18Paperclip Ai Orchestrationpaperclip-ai-orchestration is an agent skill from aradotso/trending-skills that skill for using paperclip — open-source orchestration platform for running autonomous ai-agent companies with org charts, budgets, governance, and heartbeats. # Paperclip AI Orchestration > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. Paperclip is an open-source Node.js + React platform that runs a **company made of AI agents**. It provides org charts, goal alignment, ticket-based task management, budget enforcement, heartbeat scheduling, governance, and a full audit log — so you m Developers invoke paperclip-ai-orchestration during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.4kinstalls19Cmux Terminal Multiplexercmux-terminal-multiplexer is an agent skill from aradotso/trending-skills that ai-native terminal multiplexer with programmable socket api, full playwright-equivalent browser automation, and agent team coordination — built for claude code and autonomous agent workflows. # cmux — AI-Native Terminal Multiplexer > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection cmux is a terminal multiplexer with a programmable socket API designed for AI coding agents. It provides full Playwright-equivalent browser automation, real-time terminal split management, sidebar status reporting, and agent team coordinatio Developers invoke cmux-terminal-multiplexer during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.4kinstalls20Opencli Web Automationopencli-web-automation is an agent skill from aradotso/trending-skills that turn any website into a cli using browser session reuse and ai-powered command discovery. # OpenCLI Web Automation > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. OpenCLI turns any website into a command-line interface by reusing Chrome's logged-in browser session. It supports 19 sites and 80+ commands out of the box, and lets you add new adapters via TypeScript or YAML dropped into the `clis/` folder. --- ## Ins Developers invoke opencli-web-automation during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.4kinstalls21Openmaic Classroomopenmaic-classroom is an agent skill from aradotso/trending-skills that openmaic — open multi-agent interactive classroom platform for generating immersive ai-powered learning experiences with slides, quizzes, simulations, and multi-agent discussions. # OpenMAIC — Multi-Agent Interactive Classroom > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. OpenMAIC (Open Multi-Agent Interactive Classroom) is a Next.js 16 / React 19 / TypeScript platform that converts any topic or document into a full interactive lesson. A multi-agent pipeline (LangGraph 1.1) generates slides, quizzes, Developers invoke openmaic-classroom during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.4kinstalls22Edict Multi Agent Orchestrationedict-multi-agent-orchestration is an agent skill from aradotso/trending-skills that install and use the edict (三省六部) multi-agent orchestration system with 12 specialized ai agents, real-time kanban dashboard, and audit trails. # Edict (三省六部) Multi-Agent Orchestration > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. Edict implements a 1400-year-old Tang Dynasty governance model as an AI multi-agent architecture. Twelve specialized agents form a checks-and-balances pipeline: Crown Prince (triage) → Zhongshu (planning) → Menxia (review/veto) → Shangshu Developers invoke edict-multi-agent-orchestration during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.4kinstalls23Metaclaw Evolving Agentmetaclaw-evolving-agent is an agent skill from aradotso/trending-skills that deploy and configure metaclaw — an agent that meta-learns and evolves from live conversations using skills injection, rl training, and smart scheduling. # MetaClaw Evolving Agent > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection MetaClaw is an OpenAI-compatible proxy agent that intercepts conversations, injects learned skills, and continuously improves itself through real-world interactions. It supports three modes: lightweight skills injection, immediate RL training, and a smart Developers invoke metaclaw-evolving-agent during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.4kinstalls24Freecodecamp Curriculumfreecodecamp-curriculum is an agent skill from aradotso/trending-skills that comprehensive guide for contributing to and working with freecodecamp's open-source codebase and curriculum platform. # freeCodeCamp Curriculum & Platform Development > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. freeCodeCamp.org is a free, open-source learning platform with thousands of interactive coding challenges, certifications, and a full-stack curriculum. The codebase includes a React/TypeScript frontend, Node.js/Fastify backend, and Developers invoke freecodecamp-curriculum during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments. Category AI & Agent Building with ai vertical focus supports repeatable agent-guided delivery.1.4kinstalls25Worldmonitor Intelligence DashboardWorld Monitor Intelligence Dashboard is an agent skill for building and self-hosting a real-time global intelligence dashboard from the worldmonitor TypeScript and Vite codebase. It aggregates 435 plus feeds across 15 categories, renders a dual map engine with a 3D globe.gl view and a deck.gl plus MapLibre flat map supporting 45 data layers, scores geopolitical risk across 12 signal categories per country, and tracks finance radar data from 92 exchanges. AI synthesis runs through local Ollama by default or optional Groq and OpenRouter keys, while cross-stream signal correlation detects convergence across military, economic, disaster, and escalation feeds. The project ships web, PWA, and Tauri 2 desktop targets, Vercel edge functions for feed APIs, optional Upstash Redis caching, and site variants such as tech, finance, commodity, and happy. Developers clone the repo, run npm install and npm run dev, configure .env.local for AI and map keys, register custom feeds and deck.gl layers, and deploy via Docker, Vercel, or desktop builds with troubleshooting for maps, Ollama, Redis 429s, and RTL languages.1.4kinstalls26Antigravity Managerantigravity-manager is an agent skill from aradotso/trending-skills that ai coding agent skill for antigravity manager — a tauri v2 + rust desktop app and docker service that manages multiple google/anthropic accounts and proxies them as standard openai/anthropic/gemini ap. # Antigravity Manager > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. Antigravity Manager is a professional AI account manager and proxy gateway. It takes Google (Gemini) and Anthropic (Claude) web session tokens and exposes them as standard API endpoints (OpenAI-compatible, Anthropic-native, and Gemini-native formats) with in Developers invoke antigravity-manager during operate/infra work for cloud & infrastructure tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills.1.4kinstalls27Openai Symphony Autonomous Agentsopenai-symphony-autonomous-agents is an agent skill from aradotso/trending-skills that symphony turns project work into isolated, autonomous implementation runs, allowing teams to manage work instead of supervising coding agents. # OpenAI Symphony > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. Symphony turns project work into isolated, autonomous implementation runs, allowing teams to **manage work** instead of **supervising coding agents**. Instead of watching an agent code, you define tasks (e.g. in Linear), and Symphony spawns agents that complete Developers invoke openai-symphony-autonomous-agents during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.4kinstalls28Llmfit Hardware Model Matcherllmfit-hardware-model-matcher is an agent skill from aradotso/trending-skills that terminal tool that detects your hardware and recommends which llm models will actually run well on your system. # llmfit Hardware Model Matcher > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. llmfit detects your system's RAM, CPU, and GPU then scores hundreds of LLM models across quality, speed, fit, and context dimensions — telling you exactly which models will run well on your hardware. It ships with an interactive TUI and a CLI, supp Developers invoke llmfit-hardware-model-matcher during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills.1.4kinstalls29Openclaw Configopenclaw-config is an OpenClaw Operations Runbook skill at version 3.0.0 from aradotso/trending-skills. The skill provides copy-paste bash health checks for gateway processes, JSON validation of ~/.openclaw/openclaw.json, channel policy inspection via jq, and guided fixes for channels, agents, security, and autopilot settings. Every command in the runbook is documented as tested and working. Developers reach for openclaw-config when an OpenClaw gateway stops responding, configuration JSON breaks, DM policies drift, or multi-agent channel routing needs correction. Triggers include OpenClaw setup, config inspection, and safe configuration updates across a live agent network.1.4kinstalls30Nvidia Nemoclawnvidia-nemoclaw is an agent skill from aradotso/trending-skills that nvidia nemoclaw plugin for secure sandboxed installation and orchestration of openclaw always-on ai assistants via openshell. # NVIDIA NemoClaw > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. NVIDIA NemoClaw is an open-source TypeScript CLI plugin that simplifies running [OpenClaw](https://openclaw.ai) always-on AI assistants securely. It installs and orchestrates the [NVIDIA OpenShell](https://github.com/NVIDIA/OpenShell) runtime, creates policy-enf Developers invoke nvidia-nemoclaw during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.4kinstalls31Mole Mac Cleanermole-mac-cleaner is an agent skill from aradotso/trending-skills that deep clean and optimize your mac using the mole cli tool. # Mole Mac Cleaner > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. Mole (`mo`) is an all-in-one macOS maintenance CLI that combines deep cleaning, smart app uninstallation, disk analysis, system optimization, live monitoring, and project artifact purging into a single binary. ## Installation ```bash # Via Homebrew (recommend Developers invoke mole-mac-cleaner during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments. Category AI & Agent Building with ai vertical focus supports repeatable agent-guided delivery.1.4kinstalls32Voicebox Voice Synthesisvoicebox-voice-synthesis is an agent skill from aradotso/trending-skills that expert skill for voicebox — the open-source local voice cloning and tts studio built with tauri, react, and fastapi. # Voicebox Voice Synthesis Studio > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. Voicebox is a local-first, open-source voice cloning and TTS studio — a self-hosted alternative to ElevenLabs. It runs entirely on your machine (macOS MLX/Metal, Windows/Linux CUDA, CPU fallback), exposes a REST API on `localhost:17493`, and ship Developers invoke voicebox-voice-synthesis during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.4kinstalls33Open Autoglm Phone Agentopen-autoglm-phone-agent is an agent skill from aradotso/trending-skills that expert skill for open-autoglm, an ai phone agent framework that controls android/harmonyos/ios devices via natural language using the autoglm vision-language model. # Open-AutoGLM Phone Agent > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. Open-AutoGLM is an open-source AI phone agent framework that enables natural language control of Android, HarmonyOS NEXT, and iOS devices. It uses the AutoGLM vision-language model (9B parameters) to perceive screen content and execute multi-step tasks Developers invoke open-autoglm-phone-agent during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills.1.4kinstalls34Tavily Key Generator Proxytavily-key-generator-proxy is an agent skill from aradotso/trending-skills that auto batch-register tavily api keys via browser automation and pool them behind a unified proxy gateway with web console. # Tavily Key Generator + API Proxy > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection Automates bulk Tavily account registration using Playwright + CapSolver (Cloudflare Turnstile), then pools the resulting API keys behind a unified proxy gateway with round-robin rotation, usage tracking, token management, and a web console. --- Developers invoke tavily-key-generator-proxy during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.4kinstalls35Posterskill Academic Postersposterskill-academic-posters is an agent skill from aradotso/trending-skills that ai-assisted academic conference poster generation from overleaf source using claude code. # posterskill — Academic Poster Generator > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. posterskill is a Claude Code skill that generates print-ready, interactive conference posters from your Overleaf paper source. It produces a single self-contained HTML file with a built-in drag-and-drop visual editor — no build step, no s Developers invoke posterskill-academic-posters during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.4kinstalls36Crucix Intelligence Dashboardcrucix-intelligence-dashboard is an agent skill from aradotso/trending-skills that personal intelligence agent that aggregates 27 osint data sources into a self-hosted jarvis-style dashboard with telegram/discord bots, llm analysis, and real-time alerts. # Crucix Intelligence Dashboard > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. Crucix is a self-hosted intelligence terminal that pulls from 27 open-source data feeds (satellite fire detection, flight tracking, radiation monitoring, conflict data, market prices, maritime AIS, economic indicators, and more) every 15 minutes, r Developers invoke crucix-intelligence-dashboard during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.4kinstalls37Brregbrreg is an agent skill from aradotso/trending-skills that search and retrieve norwegian company data from brønnøysundregistrene (the norwegian business registry). access all ~1.2 million registered companies in norway. # Norwegian Business Registry (Brreg) > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection Access the official open API from Brønnøysundregistrene to search and retrieve data about all registered companies in Norway. ## When to Use - When the user asks about Norwegian companies - When searching for company information in Norway - Developers invoke brreg during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.3kinstalls38Copaw Ai Assistantcopaw-ai-assistant is an agent skill from aradotso/trending-skills that personal ai assistant framework supporting multiple chat channels (dingtalk, feishu, qq, discord, etc.) with extensible skills, local/cloud deployment, and cron scheduling. # CoPaw AI Assistant Skill > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. CoPaw is a personal AI assistant framework you deploy on your own machine or in the cloud. It connects to multiple chat platforms (DingTalk, Feishu, QQ, Discord, iMessage, Telegram, Mattermost, Matrix, MQTT) through a single agent, supports custom Pytho Developers invoke copaw-ai-assistant during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.3kinstalls39Minecraftconsoles Lceminecraftconsoles-lce is an agent skill from aradotso/trending-skills that expert skill for building, running, and extending minecraftconsoles (minecraft legacy console edition) — a c++ multi-platform reimplementation with dedicated server, lan multiplayer, and modding suppo. # MinecraftConsoles (Legacy Console Edition) Skill > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. ## What This Project Is MinecraftConsoles is a C++ reimplementation/continuation of **Minecraft Legacy Console Edition v1.6.0560.0 (TU19)**, targeting modern Windows (and unofficially macOS/Linux via Wine). Goals include: - Mul Developers invoke minecraftconsoles-lce during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.3kinstalls40Picoclaw Ai Assistantpicoclaw-ai-assistant is an agent skill from aradotso/trending-skills that ultra-lightweight ai assistant in go that runs on $10 hardware with <10mb ram, supporting multiple llm providers, tools, and single-binary deployment across risc-v, arm, mips, and x86. # PicoClaw AI Assistant > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. PicoClaw is an ultra-lightweight personal AI assistant written in Go. It runs on $10 hardware with under 10MB RAM and boots in under 1 second. It supports multiple LLM providers (OpenAI-compatible, Anthropic, Volcengine), optional web search tools, and dep Developers invoke picoclaw-ai-assistant during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills.1.3kinstalls41Mirofish Offline Simulationmirofish-offline-simulation is an agent skill from aradotso/trending-skills that fully local multi-agent swarm intelligence simulation engine using neo4j + ollama for public opinion, market sentiment, and social dynamics prediction. # MiroFish-Offline Skill > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. MiroFish-Offline is a fully local multi-agent swarm intelligence engine. Feed it any document (press release, policy draft, financial report) and it generates hundreds of AI agents with unique personalities that simulate public reaction on social media — Developers invoke mirofish-offline-simulation during idea/discover work for ai & agent building tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.3kinstalls42Vinext Vite Nextjsvinext-vite-nextjs is an agent skill from aradotso/trending-skills that vite plugin that reimplements the next.js api surface for deploying anywhere, including cloudflare workers. # vinext — Next.js API on Vite, Deploy Anywhere > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. vinext is a Vite plugin that reimplements the Next.js public API surface (routing, SSR, RSC, `next/*` imports, CLI) so existing Next.js apps run on Vite instead of the Next.js compiler. It targets ~94% API coverage, supports both Pa Developers invoke vinext-vite-nextjs during build/backend work for backend & apis tasks. The skill documents triggers, prerequisites, and step-by-step workflows grounded in SKILL.md. Compatible with Claude Code, Cursor, and Codex agent runtimes that load marketplace skills. Review the Security Audits panel on this listing before installing in production environments.1.3kinstalls43Nanochat Llm TrainingThe nanochat-llm-training skill documents Karpathy's minimal nanochat harness for end-to-end LLM training on a single GPU node. It installs via uv sync from github.com/karpathy/nanochat and supports speedrun.sh to reproduce GPT-2 capability for roughly forty-eight dollars on eight H100 GPUs in about two hours. A single depth dial auto-configures width, heads, learning rate, training horizon, and weight decay for compute-optimal runs. Commands cover distributed torchrun pretraining, single-GPU base_train, quick depth-12 research iterations, CPU runcpu.sh, chat_web UI on port 8000, and chat_cli prompts. Additional scripts include scaling_laws.sh and miniseries.sh for depth sweeps. Evaluation uses DCLM CORE score with inference KV cache and ChatGPT-like web UI. Use when developers train custom LLMs with nanochat, configure depth hyperparameters, or serve a trained chat model.1.3kinstalls44Corridorkey Green ScreenThe corridorkey-green-screen skill documents CorridorKey, a neural network that solves color unmixing for green screen footage and predicts straight foreground colors plus linear alpha per pixel including motion blur and hair edges. It requires an RGB green screen frame and a coarse alpha hint mask, supports 16-bit and 32-bit EXR for VFX pipelines, and installs via uv with CUDA, MLX, or CPU extras on Windows and Linux. Windows batch installers and Linux uv sync commands download the CorridorKey checkpoint into CorridorKeyModule/checkpoints. Inference scripts process frame sequences with configurable batch size and output premultiplied or straight RGBA EXR. Optional GVM and VideoMaMa modules extend the pipeline. Use when developers need AI chroma keying, alpha matte generation, or VFX green screen compositing in Python.1.3kinstalls45Git City 3d Github VisualizationThe git-city-3d-github-visualization skill documents Git City, a Next.js 16 App Router project that renders GitHub profiles as 3D pixel art buildings with React Three Fiber and Supabase. Building height maps to contributions, width to public repos, and window brightness to stars via lib/buildings.ts metrics. Setup clones srizzon/git-city, installs dependencies, copies .env.local with Supabase URL and keys plus a GitHub personal access token, and runs dev on port 3001. Core structure spans app routes for city view, user profiles, compare mode, and admin, plus components/city for Building, CityScene, and LODManager. The skill covers adding features, customizing building decorations, achievements, GitHub API helpers, and local development workflows. Use when developers modify the 3D renderer, extend visualization features, or configure Git City locally.1.3kinstalls46Openviking Context DatabaseThe openviking-context-database skill documents OpenViking, an open-source context database for AI agents that unifies memory, resources, and skills using a filesystem paradigm instead of fragmented vector stores. It uses tiered L0, L1, and L2 context structures for hierarchical delivery, observable retrieval trajectories, and self-evolving session memory. Installation covers pip install openviking, optional ov_cli Rust CLI, and prerequisites Python 3.10+, Go 1.22+, and GCC or Clang for extensions. Configuration wires LLM providers, embedding models, and AGFS components for resource ingestion and RAG queries. Agents add memory, integrate retrieval into projects, and query context with filesystem paths. Use when developers set up OpenViking for agent memory, RAG, or skill resource management.1.3kinstalls47Json Render Generative UiThe json-render-generative-ui skill implements a generative UI framework where AI outputs JSON specs constrained by a predefined component catalog and platform renderers render them safely. Core concepts include Catalog for allowed components and actions, Spec as flat root plus elements map, Registry mapping names to implementations, and Renderer per platform. Packages cover React with shadcn, Vue, Svelte, Solid, React Native, Remotion video, react-pdf, react-email, react-three-fiber, image OG output, zustand state adapters, MCP integration, and YAML wire format. Use when developers build dynamic AI-driven UIs, stream specs, or set up json-render catalogs and registries across platforms. Agents should follow the SKILL.md workflow end to end, grounding classification in documented commands, file paths, prerequisites, and troubleshooting notes rather than improvising steps. Render AI-generated JSON UI specs into type-safe React, Vue, Svelte, and React Native components using json-render catalogs. Invoke when User mentions json-render, generative UI, AI component catalog, or streaming AI UI specs. Best for Developers building AI-driven dynamic interfaces with constrained component ca.1.3kinstalls48Karpathy Jobs Bls VisualizerThe karpathy-jobs-bls-visualizer skill supports research on Bureau of Labor Statistics Occupational Outlook Handbook data across 342 occupations. The pipeline scrapes BLS pages with non-headless Playwright, converts HTML to Markdown, extracts structured CSV fields, scores occupations via OpenRouter LLM prompts, and builds site data for a D3 treemap visualization. Developers can fork score.py with custom SYSTEM_PROMPT to create new treemap color layers such as AI exposure or robotics exposure. Key files include occupations.json, occupations.csv, scores.json, prompt.md, and site/data.json consumed by site/index.html. Use when exploring job market data, building custom LLM scoring layers, or analyzing pay, growth, and education metrics across occupations.1.3kinstalls49Modly Image To 3dThe modly-image-to-3d skill is designed for generate local GPU 3D models from images with the Modly desktop app pipeline. Modly Image-to-3D Skill > Skill by ara.so — Daily 2026 Skills collection. Modly is a local, open-source desktop application (Windows/Linux) that converts photos into 3D mesh models using AI models running entirely on your GPU — no cloud, no API keys required. Invoke when the user generates 3D models from images using local AI on GPU.1.3kinstalls50Toon FormatThe toon-format skill implements Token-Oriented Object Notation TOON, a compact human-readable JSON encoding minimizing tokens for LLM input. It combines YAML-style indentation for nested objects with CSV-style tabular layout for uniform arrays achieving roughly 40 percent token reduction while maintaining LLM comprehension. npm package @toon-format/toon provides encode and decode API plus CLI toon encode and toon decode with pipe support and --stats token comparison. Core API uses encode and decode functions converting between JSON objects and TOON strings. Use when converting JSON prompts to TOON, reducing LLM token costs, or serializing structured data for AI input with fewer tokens. Agents should follow the SKILL.md workflow end to end, grounding classification in documented commands, file paths, prerequisites, and troubleshooting notes rather than improvising steps. Encode JSON data as TOON format for ~40% fewer LLM prompt tokens with schema awareness. Invoke when User converts JSON to TOON, reduces LLM prompt tokens, or uses TOON encoding decoding. Best for Developers optimizing LLM prompts with large structured JSON payloads. Expected outcome: TOON-encoded strings or1.3kinstalls51Autoresearch GenealogyThe autoresearch-genealogy skill is designed for run autonomous genealogy research workflows with structured prompts and vault templates. A structured system of autoresearch prompts, Obsidian vault templates, archive guides, and methodology references for AI-assisted genealogy research. Built for Claude Code's autonomous research loops, adaptable to any AI tool or manual workflow. Invoke when the user explores family history, genealogy vaults, or autonomous research workflows.1.3kinstalls52Clui Cc Claude OverlayThe clui-cc-claude-overlay skill installs and configures Clui CC, a transparent floating macOS desktop overlay wrapping Claude Code CLI with multi-tab sessions, permission approval UI via PreToolUse HTTP hooks, Whisper voice input, conversation history, and a skills marketplace. Prerequisites require macOS 13 plus, Node 18 plus, Python 3.10 plus, authenticated claude CLI, and whisper-cli for voice. Installation options include app installer for non-developers or git clone dev setup with npm install and electron build. Configuration covers overlay transparency, hotkeys, permission hook endpoints, and multi-tab session management. Runs entirely local with no telemetry. Use when setting up Clui CC floating overlay, permission approval UI, or voice input for Claude Code on Mac. Agents should follow the SKILL.md workflow end to end, grounding classification in documented commands, file paths, prerequisites, and troubleshooting notes rather than improvising steps. Set up Clui CC floating macOS overlay for Claude Code with multi-tab sessions and permission UI. Invoke when User sets up clui cc, floating overlay, permission approval UI, or voice input Claude Code. Best for macOS1.3kinstalls53Translate Book ParallelThe translate-book-parallel skill is designed for translate full books (PDF/DOCX/EPUB) in parallel with resumable chunked Claude subagents. Translate Book (Parallel Subagents) > Skill by ara.so — Daily 2026 Skills collection. A Claude Code skill that translates entire books (PDF/DOCX/EPUB) into any language using parallel subagents. Invoke when the user translates entire books or large documents with parallel subagents.1.3kinstalls54Trump Code Market SignalsThe trump-code-market-signals skill is designed for analyze Trump social posts against brute-force tested rules for market signal research. Trump Code — Market Signal Analysis > Skill by ara.so — Daily 2026 Skills collection. Trump Code is an open-source system that applies brute-force computation to find statistically significant patterns between Trump's Truth Social/X posting behavior and S&P 500 movements. Invoke when the user explores Trump post analysis, market signals, or rule-based backtests.1.3kinstalls55Openclaw Rl TrainingThe openclaw-rl-training skill is designed for train personalized agents with OpenClaw-RL reinforcement learning from conversation feedback. OpenClaw-RL Training > Skill by ara.so — Daily 2026 Skills collection. OpenClaw-RL is a fully asynchronous reinforcement learning framework that converts live multi-turn conversations into training signals for personalized AI agents. Invoke when the user explores OpenClaw-RL, agent personalization, or RL from conversation feedback.1.3kinstalls56Memory Lancedb Pro OpenclawThe memory-lancedb-pro-openclaw skill is designed for configure LanceDB long-term memory for OpenClaw with hybrid retrieval and reranking. memory-lancedb-pro OpenClaw Plugin > Skill by ara.so — Daily 2026 Skills collection. memory-lancedb-pro is a production-grade long-term memory plugin for OpenClaw agents. Invoke when the user configures memory-lancedb-pro, hybrid retrieval, or OpenClaw memory scopes.1.3kinstalls57Slug Font RenderingThe slug-font-rendering skill is designed for implement GPU-accelerated Slug vector font rendering with reference HLSL shaders. Slug Font Rendering Algorithm > Skill by ara.so — Daily 2026 Skills collection. Slug is a reference implementation of the Slug font rendering algorithm — a GPU-accelerated technique for rendering vector fonts and glyphs at arbitrary scales with high quality anti-aliasing. Invoke when the user implements Slug font rendering, HLSL shaders, or GPU glyph paths.1.3kinstalls58Zeroboot Vm SandboxThe zeroboot-vm-sandbox skill is designed for spin up sub-millisecond VM sandboxes for AI agents via Zeroboot KVM forking. Zeroboot VM Sandbox > Skill by ara.so — Daily 2026 Skills collection. Zeroboot provides sub-millisecond KVM virtual machine sandboxes for AI agents using copy-on-write forking. Invoke when the user needs Zeroboot VM sandboxes, KVM forking, or agent isolation at microsecond scale.1.3kinstalls59Taiwan Md Knowledge BaseThe taiwan-md-knowledge-base skill is designed for build bilingual Taiwan knowledge base with Astro v5, D3 graph, and Markdown SSOT. Taiwan.md Knowledge Base > Skill by ara.so — Daily 2026 Skills collection. Taiwan.md is an open-source, AI-native knowledge base about Taiwan built with Astro v5. Invoke when the user builds Taiwan knowledge base, bilingual content, or D3 knowledge graph.1.2kinstalls60Grimmory Self Hosted LibraryThe grimmory-self-hosted-library skill is designed for expert knowledge for setting up, configuring, and extending Grimmory — a self-hosted book library manager supporting EPUBs, PDFs, comics, Kobo sync, OPDS, and multi-user management. Grimmory Self-Hosted Library Manager > Skill by ara.so — Daily 2026 Skills collection. Grimmory is a self-hosted application (successor to BookLore) for managing your entire book collection. Invoke when the user asks about grimmory self hosted library or related SKILL.md workflows.1.2kinstalls61Witr Process InspectorThe witr-process-inspector skill is designed for cLI and TUI tool that explains why processes, services, and ports are running by tracing causality chains across supervisors, containers, and shells. witr is a Go CLI/TUI tool that answers "why is this running?" for any process, service, or port. Invoke when the user asks about witr process inspector or related SKILL.md workflows.1.2kinstalls62Kimodo Motion DiffusionThe kimodo-motion-diffusion skill is designed for generate high-quality 3D human and humanoid robot motions using Kimodo, a kinematic motion diffusion model controlled via text prompts and kinematic constraints. Kimodo Motion Diffusion > Skill by ara.so — Daily 2026 Skills collection. Kimodo is a kinematic motion diffusion model trained on 700 hours of commercially-friendly optical mocap data. Invoke when the user asks about kimodo motion diffusion or related SKILL.md workflows.1.2kinstalls63Nightingale KaraokeThe nightingale-karaoke skill is designed for mL-powered Karaoke app in Rust using Bevy, WhisperX, and Demucs for stem separation, lyrics transcription, and pitch scoring. Nightingale Karaoke Skill > Skill by ara.so — Daily 2026 Skills collection. Nightingale is a self-contained, ML-powered karaoke application written in Rust (Bevy engine). Invoke when the user asks about nightingale karaoke or related SKILL.md workflows.1.2kinstalls64Keyid Agent Kit McpThe keyid-agent-kit-mcp skill is designed for give AI agents (Claude, Cursor) a real email address with 27 MCP tools for inbox, send, reply, contacts, search, and more via KeyID.ai. KeyID Agent Kit — MCP Email Tools for AI Agents > Skill by ara.so — Daily 2026 Skills collection. KeyID Agent Kit gives AI agents (Claude, Cursor, or any MCP client) a real, working email address with 27 tools via the Model Context Protocol. Invoke when the user asks about keyid agent kit mcp or related SKILL.md workflows.1.2kinstalls65Fabro Workflow FactoryThe fabro-workflow-factory skill is designed for skill for using Fabro, the open source AI coding workflow orchestrator that lets you define agent pipelines as Graphviz DOT graphs with human gates, multi-model routing, and. Fabro Workflow Factory > Skill by ara.so — Daily 2026 Skills collection. Fabro is an open source AI coding workflow orchestrator written in Rust. Invoke when the user asks about fabro workflow factory or related SKILL.md workflows.1.2kinstalls66Aracli Deploy ManagementThe aracli-deploy-management skill is designed for guide to deploying and managing OpenClaw-compatible AI agent systems across cloud, bare metal, and hybrid infrastructure. Deploying OpenClaw Agent Systems > Skill by ara.so — Daily 2026 Skills collection. A practical guide to deploying and managing OpenClaw-compatible AI agent systems. Invoke when the user asks about aracli deploy management or related SKILL.md workflows.1.2kinstalls67Openhanako Personal Ai AgentThe openhanako-personal-ai-agent skill is designed for build and configure personal AI agents with memory, personality, and autonomy using OpenHanako on Electron. OpenHanako Personal AI Agent > Skill by ara.so — Daily 2026 Skills collection. Multiple agents can collaborate via channel group chats or task delegation. Invoke when the user asks about openhanako personal ai agent or related SKILL.md workflows.1.2kinstalls68Codex Autoresearch LoopCodex Autoresearch is an agent skill that runs an autonomous modify→verify→keep/revert loop on codebases until a measurable goal is reached. Developers describe a single-sentence target; Codex infers scope, metrics, and verification commands, then iterates unattended with automatic git commits and reverts. Seven modes span loop optimization, planning, debugging, security audits, and release gating. Cross-run learning stores lessons; pivot protocol escalates via refinement and web search when stalled.1.2kinstalls69Code Review GraphThe code-review-graph skill build a persistent knowledge graph of your codebase so Claude reads only what matters - up to 49x fewer tokens on coding tasks. # code-review-graph > Skill by [ara.so](https://ara.so) - Daily 2026 Skills collection. `code-review-graph` builds a persistent structural map of a codebase using Tree-sitter, stores it in a local SQLite graph, and exposes it to Claude via MCP. Instead of re-reading entire projects on every task, Claude queries the graph and reads only the files in the blast radius of a change - averaging 6.8× fewer tokens on code reviews and up to 49× on daily coding tasks in large monorepos. ### pip ```bash pip install code-review-graph code-review-graph install # registers the MCP server with Claude Code ``` Requires Python 3.10+ and [uv](https://docs.astral.sh/uv/). ### Optional: semantic search support ```bash pip install code-review-graph[embeddings] ``` Enables vector embeddings via `sentence-transformers` for `semantic_search_nodes_tool`. After that, the graph updates incrementally on every file save and git commit (under 2 seconds for a 2,900-file project).1.2kinstalls70Flash Moe InferenceThe flash-moe-inference skill run 397B parameter Mixture-of-Experts LLMs on a MacBook using pure C/Metal with SSD streaming # Flash-MoE Inference Engine > Skill by [ara.so](https://ara.so) - Daily 2026 Skills collection. Flash-MoE is a pure C/Objective-C/Metal inference engine that runs **Qwen3.5-397B-A17B** (397B parameter Mixture-of-Experts) on a MacBook Pro with 48GB RAM at 4.4+ tokens/second. It streams 209GB of expert weights from NVMe SSD on demand - no Python, no ML frameworks, just C, Objective-C, and hand-tuned Metal shaders. ## Requirements - **Hardware**: Apple Silicon Mac (M3 Max or similar), 48GB+ unified memory, 1TB+ SSD with ~210GB free - **OS**: macOS 26+ (Darwin 25+) - **Tools**: Xcode Command Line Tools, Python 3.x (for weight extraction only) - **Model**: Qwen3.5-397B-A17B safetensors weights (download separately from HuggingFace) ## Installation & Build ```bash # Clone the repo git clone https://github.com/danveloper/flash-moe cd flash-moe/metal_infer # Build everything make # Verify build artifacts ls infer chat main ``` The Makefile compiles `infer.m`, `chat.m`, `main.m` with Metal shader compilation for `shaders.metal`.1.1kinstalls71Ghostling Libghostty TerminalThe ghostling-libghostty-terminal skill build minimal terminal emulators using the libghostty-vt C API with Raylib for windowing and rendering Ghostling libghostty Terminal Emulator Skill by ara so https ara so Daily 2026 Skills collection Ghostling is a minimal viable terminal emulator built on libghostty-vt the embeddable C library extracted from Ghostty https ghostty org It uses Raylib for windowing rendering and lives in a single C file The project demonstrates how to wire libghostty-vt's VT parsing terminal state and render-state API to any 2D or GPU renderer What libghostty-vt Provides VT sequence parsing SIMD-optimized Terminal state cursor styles text reflow scrollback Render state management what cells changed and how to draw them Unicode multi-codepoint grapheme handling Kitty keyboard protocol mouse tracking focus reporting Zero dependencies not even libc WASM-compatible libghostty-vt does NOT provide windowing rendering PTY management tabs splits or configuration Requirements Tool Version CMake 3 19 Ninja any C compiler clang gcc Zig 0 15 x on PATH macOS Xcode CLT or Xcode Build Run sh Clone git clone https1.1kinstalls72Weixin Agent SdkThe weixin-agent-sdk skill bridge any AI agent backend to WeChat using the weixin-agent-sdk framework with simple Agent interface login and message loop weixin-agent-sdk Skill by ara so https ara so Daily 2026 Skills collection weixin-agent-sdk is a TypeScript framework that bridges any AI backend to WeChat 微信 via the Clawbot channel It uses long-polling to receive messages no public server required and exposes a minimal Agent interface so you can plug in OpenAI Claude or any custom logic in minutes Login scan QR code once typescript import login from weixin-agent-sdk await login Credentials are persisted to openclaw run once then use start 2 Implement the Agent interface typescript import login start type Agent from weixin-agent-sdk const echo Agent async chat req return text You said req text await login await start echo Core API Agent Interface typescript interface Agent chat request ChatRequest Promise ChatResponse interface ChatRequest conversationId string Unique user conversation identifier text string Message text content media type image audio video file filePath string Local path already downloaded decrypted1.1kinstalls73724 Office Ai AgentThe 724-office-ai-agent skill documents a production AI agent system in roughly 3,500 lines of pure Python without framework dependencies. It ships 26 built-in tools, three-layer memory combining session, compressed, and vector stores, MCP and plugin support, runtime tool creation via decorators, self-repair diagnostics, and cron scheduling for recurring tasks. Installation clones the 724-office repo with minimal dependencies croniter, lancedb, and websocket-client plus optional pilk for WeChat silk audio. Configuration uses config.json for models, memory paths, and integrations including optional WeChat Work. Triggers cover self-evolving agents, runtime tool addition, MCP server connection, cron tasks, and custom decorator-based tools for 24/7 production agent systems.1kinstalls74My Brain Is Full CrewThe my-brain-is-full-crew skill installs ten specialized Claude Code agents into an Obsidian vault for knowledge management, nutrition coaching, and mental wellness support. Agents include Architect, Scribe, Sorter, Seeker, Connector, Librarian, Transcriber, Postman, Food Coach, and Wellness Guide with distinct prompts, tool restrictions, and model assignments. They coordinate through a shared Meta/agent-messages.md message board inside the vault. Setup requires Claude Code, Obsidian, git, and bash with guided onboarding for drowning knowledge workers rather than power-user optimization alone. Ten specialized agents for Obsidian vault knowledge and wellness. Roles: Architect, Scribe, Sorter, Seeker, Connector, Librarian, and more. Shared agent-messages.md coordination inside the vault. Optional Food Coach and Wellness Guide opt-in agents. Claude Code plus Obsidian second-brain installation workflow. Install 10 Obsidian vault agents for knowledge, nutrition, and wellness workflows via Claude Code.1kinstalls75Claude Peers Mcpclaude-peers-mcp is an MCP server skill by ara.so that connects multiple Claude Code instances running on one machine. A local broker daemon handles peer discovery so separate terminals or projects can list active Claude sessions and send real-time messages between them. Developers install claude-peers-mcp when coordinating multi-agent Claude Code work—splitting frontend and backend tasks, passing context between sessions, or debugging which instances are live. Trigger phrases include set up claude peers, send messages between claude sessions, and coordinate multiple claude code sessions.1kinstalls76Dingtalk Workspace Clidingtalk-workspace-cli is a Claude Code skill for DingTalk's officially open-sourced cross-platform Workspace CLI (dws). It documents how to search contacts, send messages, create calendar events and todos, and handle attendance and approvals from the command line or inside AI agent workflows. Developers reach for dingtalk-workspace-cli when DingTalk is the team collaboration hub and they need programmatic access without opening the desktop or mobile app. The skill encodes trigger phrases like automate dingtalk with dws and integrate dingtalk with AI agent so agents surface the right dws subcommands. It assumes dws is installed and authenticated on the developer machine.959installs77Polymarket Arbitrage Trading Botpolymarket-arbitrage-trading-bot is an ara.so skill from the Daily 2026 Skills collection that walks developers through building an automated dump-and-hedge arbitrage bot on Polymarket's 15-minute crypto Up/Down prediction markets. The skill covers CLOB API integration, hedge logic, and simulation mode so orders can be tested before live execution. Supported assets include BTC, ETH, SOL, and XRP across short-horizon Up/Down contracts. Trigger phrases span bot setup, dump-hedge strategy configuration, TypeScript CLOB arbitrage, and simulation runs. Developers reach for this skill when they need a structured agent playbook for Polymarket automation rather than hand-rolling order routing, position hedging, and market-selection logic from scratch. It targets engineers comfortable with trading APIs who want repeatable arbitrage workflows inside Claude Code or Cursor.945installs78Awesome Phd Cvawesome-phd-cv is a LaTeX CV and resume skill from aradotso/trending-skills in the Daily 2026 Skills collection by ara.so. It provides curated LaTeX templates for PhD students, postdocs, and researchers across three distinct use cases: ATS-optimized industry resumes for big tech, full academic CVs for faculty applications, and academic-to-industry resume conversion. Trigger phrases include faculty job application CV LaTeX, Deedy or Jake's resume templates, and ATS-safe LaTeX resumes. Developers and researchers reach for awesome-phd-cv when they need publication-ready LaTeX CV files instead of Word or Google Docs exports. The skill focuses on document generation and template selection, not job board automation or LinkedIn scraping.937installs79Web Access Claude SkillWeb Access Claude Skill from aradotso/trending-skills gives Claude Code complete internet access through a three-layer channel dispatch system, CDP-based Chrome browser automation, and parallel sub-agent task splitting for concurrent web research. Developers install it when Claude Code needs to fetch pages, scrape content, search websites, or automate a browser session instead of working from stale training data. Trigger phrases include giving Claude internet access, browsing the web, using CDP automation, and running parallel web research with sub-agents. The skill is part of the Daily 2026 Skills collection by ara.so and targets Claude Code specifically rather than generic IDE agents.932installs80Weclaw Wechat Ai Bridgeweclaw-wechat-ai-bridge is a Go integration skill that connects WeChat to AI agents including Claude, Codex, Gemini, Kimi, and OpenClaw through the WeClaw bridge. The skill covers WeClaw setup and configuration, QR-code WeChat login, inbound and outbound message routing, media conversion, and bridging WeChat traffic to OpenAI-compatible APIs. Triggers include proactive messaging, WeChat bot backends, and agent integrations where users message from WeChat instead of a web UI. WeClaw handles protocol details so agent runtimes focus on inference and tool use. Developers reach for weclaw-wechat-ai-bridge when shipping agent products to WeChat users in markets where WeChat is the primary messaging surface and a dedicated bridge is required beyond standard HTTP webhooks.923installs81Awesome Free Llm Apisawesome-free-llm-apis is a curated reference skill from ara.so's Daily 2026 Skills collection that lists LLM providers offering permanent free tiers for text inference—no trial credits and no expiry. Each listed endpoint is OpenAI SDK-compatible unless explicitly noted, with provider overviews covering rate limits, available models, and integration patterns. Developers reach for awesome-free-llm-apis when prototyping agents, swapping paid API keys for dev endpoints, or comparing which free inference APIs still work in 2026. Trigger phrases include free LLM API, free AI API key, free GPT API, and no cost LLM endpoint.919installs82Cc Connect Ai Bridgecc-connect-ai-bridge is an ara.so trending skill that bridges locally running AI coding agents—including Claude Code, Cursor, Gemini CLI, and Codex—to messaging platforms such as Feishu, Telegram, Slack, Discord, DingTalk, WeChat Work, and LINE without exposing a public endpoint. Developers reach for it when they want to chat with or schedule tasks on a local coding agent from a phone or team chat channel while keeping the agent on a private network. Triggers cover connecting Claude Code to Telegram, setting up Feishu bots, controlling agents from Slack, and sending messages to Gemini CLI from Discord. Configuration focuses on secure local bridging rather than deploying agents to cloud VMs with open ports.900installs83Bazi Skill Chinese Astrologybazi-skill-chinese-astrology is a Claude Code skill that performs interactive Four Pillars of Destiny (BaZi) chart calculation and professional destiny analysis. It walks the user through structured prompts to gather birth data, computes the complete four-pillar chart with major luck cycles and annual cycles, then produces a detailed reading that evaluates day-master strength, ten gods relationships, five-element balance, overall pattern, and provides guidance on career, relationships and health. Every analysis cross-references nine classical Chinese astrology texts to ground recommendations in traditional principles. Builders use it to embed culturally accurate Chinese astrology capabilities directly into agentic workflows without leaving the coding environment.891installs84Filmkit Fujifilm CameraFilmKit Fujifilm Camera is a Claude Code skill from aradotso/trending-skills that documents how to use FilmKit, a zero-install browser app for Fujifilm X-series cameras. The skill explains WebUSB connection setup, PTP (Picture Transfer Protocol) communication, and workflows for syncing film simulation presets and converting RAW files to JPEG directly in the browser. Developers reach for FilmKit Fujifilm Camera when building or debugging web camera integrations, testing WebUSB permissions, or automating Fuji profile management without vendor desktop tools. The SKILL.md lists eight trigger phrases such as "webusb fujifilm connection" and "filmkit preset management" so agents fire on camera-specific requests. The skill suits frontend engineers working on hardware-adjacent web features, not general photo editing pipelines.891installs85Opencli Rs Web Scraperopencli-rs-web-scraper is a Claude Code skill for the opencli-rs Rust CLI from aradotso/trending-skills. The tool pulls real-time data from 55+ websites—including Twitter, Reddit, YouTube, Bilibili, Zhihu, and Hacker News—with a single command, and can control Electron-based desktop apps such as Cursor from terminal or agent workflows. Developers reach for it when they need fast site fetchers, YAML adapter generation for new sources, or CLI integration with GitHub, Docker, and kubectl inside automated research pipelines. The skill documents install steps, trigger phrases, and adapter patterns so agents invoke opencli-rs instead of writing one-off scrapers.889installs86Vercel Labs Emulatevercel-labs-emulate wraps the vercel-labs/emulate project as a Daily 2026 Skills entry from aradotso/trending-skills. emulate spins up fully stateful, production-fidelity local HTTP servers as drop-in replacements for Vercel, GitHub, and Google APIs—designed for CI pipelines, Vitest suites, and no-network sandboxes rather than shallow mocks. Developers reach for vercel-labs-emulate when integration tests must exercise OAuth flows, deployment APIs, or GitHub endpoints without external dependencies. Triggers include emulating Vercel locally, mocking GitHub for tests, and running Google OAuth emulation in restricted environments. The skill guides agents to stand up compatible local servers so test code keeps production URL shapes while staying hermetic.884installs87Creationix Rx Data Storecreationix-rx-data-store is an expert skill from aradotso/trending-skills for RX, an embedded data store treating JSON-shaped data as a text-safe binary buffer. Developers encode a document once, then run random-access lookups directly on the encoded buffer without rebuilding an object graph or triggering GC pressure from repeated parsing. The skill covers RX encoder/decoder usage in TypeScript, converting JSON to RX, and querying encoded buffers in place—positioned as a no-SQL SQLite alternative for unstructured payloads. Use it when building services that read the same large JSON documents repeatedly and need predictable lookup latency without a full document database.873installs88Career Ops Job Searchcareer-ops-job-search is an ara.so Daily 2026 Skills wrapper for the Career-Ops pipeline built on Claude Code with 14 skill modes, a Go Bubble Tea dashboard, Playwright PDF generation, and portal scanning across 45+ company boards. YAML frontmatter lists eight trigger phrases covering setup, offer evaluation, tailored CV PDFs, portal scans, batch processing, and tracker views. The /career-ops slash commands evaluate job URLs against cv.md and config/profile.yml using a 10-dimension weighted scoring system producing A through F grades, then write reports to reports/, PDFs to output/, and rows to data/pipeline.tsv. A Go TUI offers six filter tabs and four sort modes for application status. Reach for career-ops when running a structured technical job search from the terminal—not for generic resume writing without the Career-Ops repo. Prerequisites include Node.js 18+, Go 1.21+, Playwright Chromium, and an active Claude Code session.859installs89Agent Flow Visualizationagent-flow-visualization is a skill documenting the Agent Flow VS Code extension from ara.so's Daily 2026 Skills collection. It provides real-time visualization of Claude Code agent orchestration as an interactive node graph, showing how agents make decisions, invoke tools, and branch during execution. Developers configure Claude Code hooks to feed session data into the extension for live monitoring. Trigger terms include visualizing Claude Code agents, debugging agent tool calls, tracing tool call chains, and monitoring Claude Code sessions. Reach for agent-flow-visualization when standard terminal logs are insufficient to understand multi-step agent reasoning, parallel branches, or hook-driven orchestration inside VS Code.853installs90Wewrite Wechat Ai Publishingwewrite-wechat-ai-publishing is a full-pipeline Claude Code skill from ara.so's Daily 2026 Skills collection for WeChat Official Account (公众号) articles. The workflow chains hotspot fetching, topic selection, AI drafting, SEO keyword analysis, cover image generation, WeChat-specific formatting, and draft-box publishing in one agent invocation. Eight Chinese-language triggers cover requests like writing a 公众号 article, fetching hotspots, generating cover images, and publishing to the WeChat draft box. Developers reach for wewrite-wechat-ai-publishing when they need repeatable, agent-driven WeChat content production without manually switching between research, writing, SEO, design, and publishing tools.851installs91Tg Ws Proxy Telegram Socks5tg-ws-proxy-telegram-socks5 is a CLI skill from ara.so Daily 2026 Skills for developers whose Telegram Desktop clients cannot reach Telegram DCs over direct MTProto connections. TG WS Proxy runs a local SOCKS5 server that reroutes Telegram traffic through WebSocket (WSS) tunnels to Telegram data centers, restoring connectivity in restricted networks. The skill covers installation from source, Python-based setup, SOCKS5 configuration inside Telegram Desktop, and troubleshooting when a standard Telegram proxy fails. Its frontmatter lists eight trigger phrases including tg-ws-proxy installation, websocket bypass, and DC websocket tunnel setup. Reach for it when MTProto is filtered but HTTPS/WebSocket paths remain open—not when you only need a generic HTTP proxy for browsers. Step-by-step guidance targets developers who must run tg-ws-proxy from source, bind a local SOCKS5 listener, and point Telegram Desktop proxy settings at the tunnel endpoint on restrictive corporate or regional networks.850installs92Microsoft Rust Trainingmicrosoft-rust-training is a Claude Code skill bundling seven structured Rust training books maintained by Microsoft, covering Rust from multiple entry points including C/C++, C#/Java, and Python backgrounds. The curriculum spans beginner to expert levels with exercises, diagrams, and playgrounds across topics like async Rust, type-driven correctness, and engineering practices. Developers reach for microsoft-rust-training when learning Rust from scratch, switching from another language, or following the Microsoft Rust curriculum for structured progression rather than scattered documentation.849installs93Jkvideo Bilibili React Nativejkvideo-bilibili-react-native is an expert agent skill for building and extending JKVideo, a React Native Bilibili-like client. It guides DASH adaptive playback with multiple quality levels in Expo, danmaku bullet comments, WBI signature handling for Bilibili API calls, WebSocket-based live streaming, TypeScript API integration, and a download manager with LAN sharing. Developers reach for jkvideo-bilibili-react-native when triggers include building a Bilibili React Native app, adding DASH playback to Expo, implementing danmaku, or wiring bilibili API integration. The skill fits teams shipping third-party video clients that mirror Bilibili UX patterns without reimplementing protocol and player edge cases from scratch.847installs94Ai Scanner Garakai-scanner-garak wraps AI Scanner (0din-ai/ai-scanner), an open-source Ruby on Rails web application for AI model security assessments built on NVIDIA garak. The skill configures scans with 179 security probes spanning 35 vulnerability families, including OWASP Top 10 LLM checks, recurring scheduled scans, and PDF or SIEM result exports. Developers reach for ai-scanner-garak before exposing chatbots, agents, or model APIs to end users when manual red-teaming cannot cover the full probe matrix. Triggers include running AI safety assessments, garak penetration testing setup, and scheduling recurring LLM security scans.846installs95Mcp Brasil Public Apismcp-brasil-public-apis is a Claude Code skill by ara.so that installs and configures an MCP server bridging agents to 28 Brazilian public APIs spanning economy, legislation, transparency, judiciary, elections, environment, and health. Developers use it when they want to query IBGE, Banco Central, legislative records, or transparency portals from an agent without writing bespoke API clients for each source. Triggers include phrases like "use mcp-brasil," "query Brazilian public data," and "setup mcp server for Brazil APIs." The skill follows the Model Context Protocol so tools appear natively inside supported coding agents.835installs96Holyclaude Ai WorkstationHolyClaude AI Workstation is a Claude skill from aradotso's trending-skills collection that deploys a containerized AI development workstation via Docker Compose. The container bundles Claude Code CLI, a CloudCLI web UI, Chromium headless browser with Playwright, and five AI model CLIs in one environment. Developers reach for HolyClaude when setting up Claude Code in Docker, configuring a web UI for coding agents, or running headless browser automation alongside multiple model CLIs. Triggers include HolyClaude troubleshooting, environment configuration, and docker compose deployment workflows.832installs97Type4me Macos Voice Inputtype4me-macos-voice-input is a development skill for engineers customizing Type4Me, a Swift macOS voice-input tool that captures speech through a global hotkey and inserts transcribed text into any application. The skill covers adding ASR providers, implementing the SpeechRecognizer protocol, configuring local recognition with Sherpa, wiring Volcengine cloud speech APIs, and enabling custom prompt modes that run LLM text optimization on transcripts. It also documents building and deploying Type4Me from source and troubleshooting when voice input fails. All storage stays local by default, making the skill relevant when you need privacy-preserving dictation or a branded voice layer inside internal macOS tooling. Reach for it when extending Type4Me with a new cloud ASR service, tuning local offline models, or debugging hotkey capture pipelines—not when building iOS or cross-platform mobile dictation from scratch without the Type4Me codebase.830installs98Obliteratus Abliterationobliteratus-abliteration is an advanced agent skill for machine learning engineers researching transformer refusal geometry and modifying open-source language models with the OBLITERATUS toolkit. The skill documents one-click model liberation workflows: extracting refusal directions from transformer activations, analyzing refusal geometry, and applying surgical abliteration to Hugging Face model weights so outputs no longer trigger built-in refusal behaviors. Triggers include abliterating a model, removing refusal from an LLM, running abliteration on a Hugging Face checkpoint, and using OBLITERATUS to analyze guardrail directions. Reach for it when you own the model weights, run controlled red-team or alignment research, and need reproducible abliteration steps rather than ad-hoc weight edits. The skill is not a general fine-tuning guide—it focuses narrowly on refusal-direction extraction and abliteration mechanics for uncensored research variants in controlled lab environments with documented checkpoints.829installs99Markit Markdown Convertermarkit-markdown-converter is an ara.so Daily 2026 Skills entry for converting almost any file or URL to markdown via the markit-ai CLI and pluggable TypeScript SDK. Supported inputs include PDF, DOCX, PowerPoint, Excel, HTML, EPUB, Jupyter notebooks, RSS feeds, CSV, JSON, YAML, images with EXIF, and audio with AI transcription. Developers reach for markit-markdown-converter when ingesting legacy documents into RAG pipelines, writing markit plugins, installing the markit CLI, or embedding markit as a library in TypeScript projects. Triggers cover convert PDF to markdown, extract markdown from URLs, and convert images or audio with AI assistance.824installs100Legalize Es Spanish Legislationlegalize-es-spanish-legislation is an ara.so skill for working with the legalize-es Git repository containing 8,600+ Spanish laws as Markdown files with full Git history of legislative reforms. Agents can search Spanish law, browse BOE legislation, locate articles, compare versions, and track regulatory changes using repository triggers such as query BOE law or compare versions of Spanish law. Developers reach for legalize-es-spanish-legislation when building Spain-compliant products, drafting policy features, or auditing which legal articles changed over time. The skill navigates a legislation corpus rather than generating application payment or frontend UI code.821installs101Flipoff Split Flap Displayflipoff-split-flap-display is an expert agent skill for the FlipOff split-flap display emulator—a free, offline-capable web app that turns any browser or TV into a retro airport departure board. The skill covers embedding the emulator on landing pages and internal dashboards, customizing messages and quotes, adjusting grid size and colors, configuring sound and flip animations, and adding keyboard controls. Developers reach for flipoff-split-flap-display when they want a distinctive retro UI element without building split-flap physics, timing, and styling from scratch.818installs102Mac Code Local Ai Agentmac-code-local-ai-agent is an ara.so trending-skills module for running mac-code, a CLI AI coding agent on Apple Silicon at $0/month. The skill guides loading a 35B reasoning model locally through llama.cpp or MLX with web search, shell, and file tools attached. Triggers include local LLM agent setup, Claude Code alternatives on Mac, and out-of-RAM inference tuning on 16GB machines. Developers reach for it when they want an on-device coding agent with tool use instead of cloud API billing. The skill covers setup, usage, and memory constraints for local inference.811installs103K Skill Korean Ai Toolsk-skill-korean-ai-tools is a Korean service automation skill pack for Claude Code, Codex, and Cursor agents published in ara.so's Daily 2026 Skills collection. The suite exposes triggers and install commands for SRT/KTX train booking, KBO results, lottery number checks, KakaoTalk messaging, Seoul subway arrival info, HWP document conversion, and postal code lookup. Install the full pack with npx k-skill install or select individual modules such as npx k-skill install srt for train booking only. Developers reach for k-skill when building Korea-local agent workflows that must interact with domestic APIs and file formats unavailable in generic international skill libraries.810installs104Codex Plugin Cccodex-plugin-cc is an ara.so Daily 2026 Skills entry that connects OpenAI Codex to Claude Code for code reviews, adversarial reviews, bug investigations, and background delegated tasks. Trigger phrases include review my code with Codex, delegate a task to Codex, and install codex plugin claude code, plus commands to check background Codex job status. Developers use codex-plugin-cc when they want a second model reviewing or exploring code without leaving Claude Code. The skill focuses on plugin setup, review invocation, and monitoring delegated jobs rather than replacing either agent's native editing workflow.800installs105Pump Analyzer Solanapump-analyzer-solana is a Claude Code skill from aradotso/trending-skills that scaffolds PumpAnalyzer, a static front-end platform for real-time Pump.fun token monitoring on Solana. The implementation uses pure HTML, CSS, and JavaScript with no build tools, connecting to Pump.fun WebSocket feeds for live launch tracking, analytics, and alerts. Developers reach for pump-analyzer-solana when setting up memecoin dashboards, integrating Pump.fun websocket data, adding real-time token launch alerts, or connecting a Solana wallet to a tracker interface. Trigger phrases include track pump.fun tokens, monitor solana token launches, and build memecoin dashboard. The skill produces an instantly usable monitoring UI rather than requiring a Node or React toolchain.800installs106Caveman Token Optimizercaveman-token-optimizer is an agent-behavior skill from aradotso/trending-skills that rewrites Claude Code and Codex replies into terse caveman-speak, reporting about 65% average output token reduction and up to 87% in favorable cases while preserving technical precision. Install triggers include phrases like "install caveman mode" and "reduce token usage caveman." Developers reach for caveman-token-optimizer when long agent transcripts inflate API bills or slow loops, but they still need correct commands, code, and debugging guidance in responses.792installs107Turboquant PytorchTurboQuant PyTorch is a Claude Code skill with a from-scratch PyTorch implementation of Google's TurboQuant method from ICLR 2026 for LLM key-value cache compression. It applies two-stage vector quantization—random rotation, Lloyd-Max quantization, and QJL residual correction—to shrink KV cache footprint while preserving attention inner products for long-context inference. Developers reach for TurboQuant PyTorch when optimizing transformer serving memory, experimenting with 3-bit KV cache compression, or reproducing TurboQuant research in production PyTorch stacks. Triggers include compress KV cache, quantize key value cache, reduce KV cache memory, and apply QJL residual correction.782installs108App Store Screenshots Generatorapp-store-screenshots-generator is a skill from aradotso/trending-skills in the Daily 2026 Skills collection that builds production-ready App Store screenshots for iOS apps. The workflow uses AI coding agents with Next.js and html-to-image to render and export marketing screenshot assets instead of manual design tooling. Developers reach for app-store-screenshots-generator when preparing App Store Connect submissions, refreshing promotional images, or automating screenshot pipelines for iPhone apps. Triggers include requests to build App Store screenshots, generate marketing images, or automate promotional asset creation. The skill targets Claude Code, Cursor, Windsurf, and Codex agents that can scaffold a Next.js screenshot generator and export PNG assets ready for App Store listing requirements.779installs109Open Pencil Design Editoropen-pencil-design-editor is a skill from aradotso/trending-skills documenting OpenPencil, an open-source AI-native design editor in the Daily 2026 Skills collection. OpenPencil reads and writes native Figma .fig files, exposes a headless CLI, ships an MCP server for agent access, and provides a Vue SDK for custom interfaces. Developers reach for open-pencil-design-editor when automating design file edits, building self-hosted design tooling, or scripting Figma-compatible assets without the proprietary Figma desktop app. Triggers include opening .fig files programmatically, running design-editor CLI commands, or wiring an MCP server for design automation. The skill suits teams integrating design files into CI pipelines, agent workflows, or Vue-based internal design tools.773installs110Tribev2 Brain Encodingtribev2-brain-encoding is an aradotso trending-skills agent skill for Meta's TRIBE v2 foundation model that predicts cortical fMRI responses to naturalistic video, audio, and text inputs. The model combines LLaMA-based language understanding with multimodal encoders for in-silico neuroscience experiments without collecting new neuroimaging data. Developers reach for tribev2-brain-encoding when building brain encoding pipelines, running TRIBE v2 inference on stimulus sets, or exploring multimodal predictors of cortical activity. Trigger phrases include fMRI encoding models, brain response prediction from video, and tribev2 training or inference workflows for computational neuroscience research.773installs111Pgmicro Postgres Sqlitepgmicro-postgres-sqlite is a skill from aradotso/trending-skills documenting pgmicro in the Daily 2026 Skills collection. pgmicro is an in-process reimplementation of PostgreSQL backed by SQLite-compatible storage, parsing PostgreSQL SQL with the real PostgreSQL parser libpg_query. Developers reach for pgmicro-postgres-sqlite when they need PostgreSQL semantics in embedded apps, local tooling, or serverless functions without provisioning a Postgres daemon. Triggers include embedding Postgres in an app, running PostgreSQL without a server, or setting up the pgmicro JavaScript SDK or Rust embeddable library. The skill suits mobile clients, CLI tools, and edge deployments where a separate database process is costly or impossible while PostgreSQL SQL compatibility remains required.772installs112Polymarket Arbitrage Botpolymarket-arbitrage-bot is a TypeScript Claude Code skill from aradotso’s trending-skills collection that automates dump-and-hedge arbitrage on Polymarket’s 15-minute Up/Down prediction markets. It wires CLOB order execution, strategy configuration, and a simulation mode so developers can test fills before live trading. Trigger phrases cover bot setup, hedge strategy tuning, and CLOB TypeScript integration. Reach for this skill when implementing automated Polymarket market-making or arbitrage bots rather than writing CLOB client boilerplate from scratch. The skill encodes strategy parameters and execution flow for Up/Down short-window markets.767installs113Polymarket Copy Trading Botpolymarket-copy-trading-bot is an ara.so trending-skills entry that scaffolds a TypeScript bot to monitor a target Polymarket wallet and mirror BUY trades into the developer's own account. The skill covers Polymarket CLOB API integration on Polygon, websocket or polling trade detection, environment setup, and automated order placement triggered by the watched wallet. Documented triggers include copy-trade setup, CLOB order bot configuration, and websocket trade monitoring requests. Developers reach for polymarket-copy-trading-bot when automating prediction-market follow trading instead of manually replicating each BUY on Polymarket.762installs114Taiwan Equity Research Coveragetaiwan-equity-research-coverage (My-TW-Coverage) is an aradotso/trending-skills database covering 1,735 Taiwan-listed companies with wikilink knowledge graphs, supply chain mapping, and financial data tools. Developers and research engineers query buzzword themes, refresh valuation multiples, audit report quality, and build thematic investment screens without manually collating exchange filings. Triggers include adding tickers, updating financials, and generating ecosystem supply-chain graphs for Taiwan equities. The skill fits agent workflows that turn fragmented TWSE data into linked research artifacts and repeatable screens.761installs115Torchcode Pytorch Interview Practicetorchcode-pytorch-interview-practice is an ara.so Daily 2026 Skills entry for TorchCode, a self-hosted, Jupyter-based LeetCode-style environment for PyTorch coding interviews. Triggers include implementing PyTorch operators from scratch, practicing softmax, LayerNorm, attention, transformer components, and running the TorchCode judge for automated feedback. Developers reach for torchcode-pytorch-interview-practice when preparing for ML engineering interviews, validating low-level tensor implementations, or drilling GPT-2 building blocks without relying on high-level nn.Module shortcuts.758installs116Codex Session Patchercodex-session-patcher is an aradotso/trending-skills agent skill backed by a lightweight Python utility for AI coding session files. It detects refusal responses in Codex CLI, Claude Code, and OpenCode session transcripts, removes or rewrites those segments, and can inject CTF or pentest-oriented context prompts so sessions resume cooperatively. Security researchers and developers hit hard refusals mid-session reach for codex-session-patcher instead of manually editing JSON session stores. Triggers include cleaning codex session refusals, patching AI refusal responses, and resuming codex sessions after a block. Use only on sessions you own and within policy-bound security research contexts.756installs117Learn Docker K8s Gamelearn-docker-k8s-game is an open-source interactive learning skill from aradotso/trending-skills that teaches Docker, Linux, networking, and Kubernetes via the nocappuccino story-driven game inside Claude Code or compatible editors. Triggers include "let's play the docker learning game" and "start the kubernetes learning adventure." Developers reach for learn-docker-k8s-game when they want hands-on container and orchestration fundamentals without leaving the agent session, especially before touching production Dockerfiles, compose files, or cluster manifests.750installs118Claude Better Cliclaude-better-cli is a performance-oriented CLI skill from aradotso/trending-skills for a compatibility-first reimplementation of the official Claude CLI called claude-better. It advertises up to 73% faster startup and up to 80% lower resident memory while preserving drop-in command compatibility for existing scripts and habits. Developers reach for claude-better-cli when terminal harnesses feel sluggish, CI agents spend time waiting on CLI boot, or memory pressure appears during repeated Claude invocations.746installs119Ministack Aws Emulatorministack-aws-emulator is a skill from aradotso/trending-skills for MiniStack, a free MIT-licensed local AWS emulator positioned as a LocalStack replacement. MiniStack emulates 25+ AWS services on a single port with no AWS account, credentials, or license required. The skill helps developers set up local endpoints for boto3 code, run DynamoDB, S3, and Lambda locally, mock AWS in CI/CD pipelines, and replace paid LocalStack setups. Trigger phrases include emulate AWS services locally, test boto3 without AWS, and local AWS endpoint for CI/CD. Reach for ministack-aws-emulator when integration tests need AWS APIs without cloud cost or credentials.745installs120Wildworld Datasetwildworld-dataset is an agentic coding skill from aradotso/trending-skills for working with the WildWorld large-scale action-conditioned world modeling dataset captured from a photorealistic ARPG game. The dataset includes 108M+ frames with per-frame annotations, 450+ distinct actions, and explicit state information designed for generative world modeling and WildBench benchmark evaluation. ML engineers and research developers reach for wildworld-dataset when loading ARPG gameplay sequences, inspecting frame-level annotations, or prototyping action-conditioned video models inside coding agent sessions. The skill supports triggers around WildWorld annotations, action-conditioned video experiments, and benchmark evaluation workflows without requiring manual dataset documentation lookups each session.741installs121Phantom Ai Coworkerphantom-ai-coworker is a Claude Agent SDK skill for deploying Phantom, an AI coworker that runs on a dedicated machine with session-spanning memory, self-evolution, dynamic MCP tool creation, and Slack or email identity. Unlike stateless chatbots, Phantom registers its own MCP servers, persists context across runs, and operates through configured channels on Docker or bare metal. Developers reach for phantom-ai-coworker when wiring long-running agent infrastructure—memory stores, tool registries, channel bots, and deployment manifests—rather than one-off prompt chains. Triggers include phantom MCP server setup, Slack bot agent configuration, persistent memory agents, and Docker-based Phantom deployment from the skill manifest.728installs122Kordoc Korean Document ParserKordoc Korean Document Parser is a document ingestion skill from aradotso/trending-skills built around kordoc, a TypeScript library and CLI for parsing Korean government documents including HWP 5.x, HWPX, and PDF into Markdown and structured data. Triggers cover parse hwp file to markdown, convert korean document to text, extract text from hwpx, compare two hwp documents, extract form fields, and set up kordoc mcp server. The skill supports CLI usage, programmatic API integration, and MCP server setup for agent and RAG pipeline workflows. Developers reach for Kordoc Korean Document Parser when building Korean-language document automation that must handle legacy HWP formats beyond standard PDF text extraction. Output targets clean Markdown suitable for downstream indexing and agent consumption.726installs123Claw Code Harnessclaw-code-harness is a skill from aradotso/trending-skills for Claw Code—a clean-room Python rewrite of the Claude Code agent harness with an in-progress Rust port. The skill exposes CLI workflows for manifest inspection, parity auditing, tool and command inventory, subsystem listing, and tool port metadata review. Documented triggers include phrases like set up claw-code harness, run parity audit claw-code, and claw-code manifest summary. Developers reach for it when comparing harness behavior, auditing tool coverage, or exploring subsystem boundaries without spelunking minified upstream internals. The Daily 2026 Skills collection positions it for engineers porting, testing, or documenting agent harness capabilities from ara.so. Use it during agent platform work when inventory accuracy and parity evidence matter before shipping custom tool integrations.720installs124Clawd Code Python Portclawd-code-python-port is a trending skill from aradotso that guides developers through an independent Python port of the Claude Code agent harness. It documents how tools register and dispatch, how commands extend the CLI, and how task orchestration runs multi-step agent jobs via the oh-my-codex entrypoint. Triggers include adding tools, extending commands, running parity audits, and verifying the Python workspace. Developers reach for it when they need a local, hackable agent runtime in Python rather than only using the bundled TypeScript Claude Code client. The skill supports parity verification against the original harness behavior.713installs125See Through Anime Layer Decompositionsee-through-anime-layer-decomposition is a Claude Code skill from aradotso/trending-skills that teaches agents how to run See-through, a diffusion-based framework for splitting one anime illustration into semantic body-part layers exported as a fully editable PSD. The skill encodes triggers such as decompose anime character into layers, split anime illustration into PSD layers, and generate layered PSD from anime image so an agent selects the right decomposition workflow instead of generic image editing. Developers reach for see-through-anime-layer-decomposition when a single character render must become separated hair, face, torso, limbs, and accessories for 2.5D parallax, Live2D-style rigs, or compositing in Photoshop. The workflow targets anime-style art with see-through layer separation rather than photoreal cutouts. Outputs are layered PSD files ready for timeline animation, puppet warp, or engine import.703installs126Open Agent Sdkopen-agent-sdk is a Claude Code skill from aradotso/trending-skills that teaches developers to build and deploy autonomous AI agents in-process using Open Agent SDK, an open-source alternative to @anthropic-ai/claude-agent-sdk that runs without CLI dependencies. The skill covers TypeScript and Node.js integration patterns for serverless functions, cloud workers, and backend services where spawning a separate agent CLI process is impractical. Developers reach for open-agent-sdk when replacing claude-agent-sdk with an open-source in-process runtime, running agent loops inside existing Node servers, or deploying agents to serverless platforms. Trigger phrases include open-agent-sdk setup, deploy claude agent without CLI, and run AI agent loop in process. The skill walks through package installation, agent loop configuration, and deployment patterns for @shipany/open-agent-sdk so agents execute entirely within the host application process.698installs127Emdash Cmsemdash-cms is an agent skill for building with EmDash, a full-stack TypeScript CMS built on Astro and Cloudflare positioned as an extensible WordPress successor with a plugin system. The skill covers setup on Astro sites, content type creation, plugin development, content collection queries, WordPress migration, Cloudflare D1 configuration, and authentication. Triggers include phrases like set up EmDash CMS, add EmDash to my Astro site, and configure EmDash with Cloudflare D1. Developers reach for emdash-cms when they want a TypeScript-native CMS with Cloudflare edge deployment instead of traditional PHP WordPress stacks.665installs128Openclaude Multi Llmopenclaude-multi-llm is an agent skill for OpenClaude, a fork of Claude Code that routes all LLM calls through an OpenAI-compatible API via environment variable configuration. The skill supports GPT-4o, DeepSeek, Gemini, Ollama, Groq, and 200+ models while preserving Claude Code's full tool system. Triggers include use Claude Code with OpenAI, run Claude Code with a different model, and connect Gemini to Claude Code. Developers reach for openclaude-multi-llm when they want Claude Code's agent tools and workflows but need to swap the underlying model provider for cost, locality, or capability reasons.665installs129Cc Gateway Ai Proxycc-gateway-ai-proxy is a skill from aradotso/trending-skills for deploying CC Gateway, a TypeScript reverse proxy that sits between Claude Code clients and the Anthropic API. The gateway normalizes device fingerprints, rewrites device IDs, and controls telemetry to provide privacy-preserving API proxying across multiple Claude Code machines. Developers use it with Docker setup workflows when they need centralized proxy control instead of direct client-to-API connections. Trigger phrases include cc-gateway docker setup, proxy claude code through a gateway, and configure anthropic api reverse proxy.661installs130Open Multi Agent Orchestrationopen-multi-agent-orchestration is a skill from aradotso/trending-skills for the open-multi-agent TypeScript framework. It teaches agents to stand up production-grade multi-agent teams with dependency-graph task scheduling, inter-agent communication, shared memory, parallel execution, custom tools, and mixed Claude plus GPT models in one workflow. Developers reach for open-multi-agent-orchestration when building agent pipelines that must run tasks in parallel, respect dependencies, or combine models rather than a single prompt chain. Triggers include setting up multi-agent teams, orchestrating collaborators, and defining task pipelines with explicit DAG edges.657installs131Claude Code Source Recoveryclaude-code-source-recovery is an ara.so Daily 2026 Skills entry for studying recovered Claude Code 2.1.88 TypeScript source code. The skill explains CLI architecture, the command system, MCP integration points, and Ink/React terminal UI components recovered from source maps. Developers use it to explore how a production AI coding CLI wires sessions, commands, MCP servers, and terminal rendering when building or auditing comparable agent tooling. Triggers include studying command implementation, MCP integration source, terminal UI components, and codebase structure after source map recovery. Reach for this skill when reverse-engineering agent CLI patterns from Claude Code internals rather than writing unrelated application features.640installs132Pretext Text Measurementpretext-text-measurement is a skill for developers building canvas UIs, rich-text editors, and custom layout engines in JavaScript or TypeScript. Pretext measures multiline, variable-width, and rich text without calling getBoundingClientRect or offsetWidth, avoiding expensive DOM reflows during layout passes. The skill triggers on tasks like calculating paragraph height, shrink-wrapping text width, wrapping text on canvas, or laying text around images. It documents how to use Pretext for accurate sizing when the browser DOM is too slow or unavailable, such as in WebGL overlays, PDF-like renderers, or performance-sensitive dashboards. Developers reach for this skill when standard DOM measurement APIs cause jank or when layout must run entirely off-DOM while still matching real font metrics.640installs133Sheets Terminal Spreadsheetsheets-terminal-spreadsheet is an aradotso/trending-skills guide for the Sheets terminal spreadsheet TUI by maaslalani, written in Go for viewing and editing CSV files directly in the terminal. Documented triggers include edit CSV in terminal, sheets TUI spreadsheet, vim keybindings spreadsheet, and open CSV file in terminal. The skill targets developers who live in vim-style TUIs and need quick budget or operations CSV edits without launching Excel or Google Sheets. Agents invoke it when users request terminal spreadsheet workflows or mention the maaslalani sheets tool. The workflow emphasizes in-terminal navigation and cell editing for plain CSV artifacts common in ops and finance tracking.623installs134Omnivoice TtsOmniVoice TTS is an expert agent skill that turns text into high-quality speech across more than 600 languages using a state-of-the-art diffusion language model. It enables instant voice cloning from just a few seconds of reference audio, text-driven voice design, and efficient batch generation. Designed for Python-based agents and scripts, it gives solo builders the ability to add realistic multilingual voice output to products, prototypes, accessibility features, or interactive experiences without needing large training datasets or complex pipelines.598installs135Metatron Pentest AssistantMETATRON is a fully local CLI-based AI penetration testing assistant designed for solo builders and indie developers who need to run professional-grade security scans without sending data to any cloud service. It combines industry-standard reconnaissance tools with a locally hosted fine-tuned LLM that interprets results, suggests exploits, and recommends fixes. All findings are stored in a MariaDB database for historical tracking, trend analysis, and report generation. The skill runs on Parrot OS Linux, uses Ollama to power its metatron-qwen model, and includes modules for database management, tool orchestration, DuckDuckGo-based CVE lookup, and extensible agentic workflows. Perfect for developers who want to integrate automated security testing into their local development and pre-launch routines.584installs136Tong Jincheng Relationship SkillTong Jincheng Relationship Skill is an agent skill that channels the unfiltered, straight-talking worldview of 童锦程 (Deep Affection Patriarch) to analyze interpersonal, romantic and life-choice problems. Built from 200,000 words of original material, it applies nine core heuristics such as "Uncertainty means she doesn't like you", "Human nature cannot be tested", "Attraction beats pleasing" and "Give people a ladder they can climb". Instead of repeating slogans or soft advice, it forces clarity on unspoken needs, exposes insecurity behind bragging, and distinguishes pre-success from post-success realities. Solo builders activate it in Claude Code with any of the listed Chinese or English trigger phrases and receive concise, actionable reframes that cut through self-deception on both personal and professional fronts.575installs137Parlor On Device AiParlor On-Device AI is a real-time multimodal voice and vision assistant that runs 100% locally on your hardware. It uses Gemma 4 E2B through LiteRT-LM to process speech and camera input while Kokoro TTS generates natural voice replies. The stack includes a browser client capturing microphone and camera, a FastAPI WebSocket server for bidirectional streaming, Silero VAD for hands-free activation, and barge-in support so users can interrupt the assistant mid-response. No API keys, no recurring costs, and no data leaves your machine. Developers use it to create private, responsive AI companions, local voice interfaces, or on-device multimodal prototypes that feel instantaneous because everything stays on the GPU and CPU you already own.565installs138Free Code Claude Clifree-code is an open-source fork of Anthropic's Claude Code CLI that gives developers a terminal-native AI coding agent without restrictions. It eliminates telemetry, removes injected guardrails, and unlocks every experimental feature flag disabled in the official npm release. The skill provides a one-liner installer that handles Bun, clones the repository, builds with full experimental support, and places the free-code binary on your PATH. It works with Anthropic, OpenAI Codex, AWS Bedrock, Google Vertex AI, and Anthropic Foundry so you can switch providers instantly. Developers use it to gain complete control over their local coding agent, run advanced reasoning modes, and avoid any vendor lock-in or data collection.560installs139Paper2code Arxiv Implementationpaper2code-arxiv-implementation is an agent skill that reads any arXiv paper URL and produces a working Python implementation where every implementation choice is explicitly linked to the originating section, equation or figure in the original paper. Instead of letting the model silently fill knowledge gaps, the skill surfaces every ambiguity, missing detail or underspecified component in a separate audit report so the developer can address them consciously. It supports common deep-learning frameworks via command flags and works inside Claude Code, Cursor and similar agents. Developers use it when they want to reproduce, extend or productionize a machine-learning method without introducing hidden assumptions. The result is reproducible code that stays faithful to the source material and an audit trail that makes future updates or audits straightforward.560installs140Fieldtheory Cli BookmarksField Theory CLI (ft) is a command-line tool that downloads, indexes, classifies, and exposes your X/Twitter bookmarks for use by local AI coding agents. It works by reading your Chrome session on macOS or using OAuth on other platforms, eliminating the need for Twitter API keys. Once synced, you can instantly search thousands of saved posts with natural language, run LLM-powered classification, and let agents like Claude Code pull relevant bookmarks as context during development, research, or content work. The tool is lightweight, stores everything locally, and turns passive bookmarking into an active, queryable second brain that grows with you.559installs141Awesome Design Md JpAwesome Design MD JP is a curated collection of DESIGN.md files that teach AI coding agents how to produce accurate Japanese user interfaces. It extends the Google Stitch DESIGN.md standard with culturally correct CJK typography rules, Japanese font stacks, appropriate line heights, letter spacing, kinsoku shori punctuation handling, OpenType font features, and mixed Japanese-Western typesetting. developers use it when creating Japanese-facing web apps, SaaS products, or any interface where visual accuracy and typographic harmony matter to users and to the AI agents writing the frontend code. The skill removes guesswork and cultural mismatches that commonly appear when Western-centric prompts are sent to Claude, Cursor, or Codex.557installs142Tong Jincheng Interpersonal Skilltong-jincheng-interpersonal-skill is an agent skill that equips your AI coding agent with a distilled cognitive framework from 童锦程 (Tong Jincheng). It draws from nine core mental models such as 'Uncertainty is Rejection', 'Attraction over Pleasing', 'Give a Ladder', and 'Human Nature Cannot Be Tested'. The skill lets the agent analyze real personal situations involving relationships, attraction, decision-making, and self-understanding with the same direct, experience-based lens as the source material. It is built from approximately 200,000 words of original transcripts and writings rather than surface-level quotes. Once installed, simply invoke it with trigger phrases and the agent will return clear, actionable perspective instead of generic advice or theatrical role-play.544installs143Claude Code Source StudyClaude Code Source Study is a 25-article series that systematically dissects Anthropic’s Claude Code CLI. Each article walks through specific modules with precise file references, line numbers, and code excerpts so you can understand exactly how a production-grade AI coding agent is built. The series covers system prompt design, multi-agent coordination, tool-use infrastructure, permission models, terminal UI with Ink, state management, and startup optimizations. Written in Chinese, it translates real-world Anthropic engineering decisions into reusable patterns that developers and small teams can apply when creating their own AI agents, CLIs, or autonomous coding tools. Perfect when you want to move beyond prompt hacking and study battle-tested architecture from a shipped product.539installs144Cabinet Ai Knowledge BaseCabinet AI Knowledge Base documents how to install and run Cabinet, an AI-first startup OS from the ara.so Daily 2026 collection where notes, agents, and apps live as markdown on disk. Developers use it when they want a self-hosted alternative to Notion-plus-chat silos: optional KB_PASSWORD auth, DOMAIN configuration, Claude Code CLI as a prerequisite, and dev:all serving the web UI with wizard-driven team setup. The skill covers quick start via npx create-cabinet, manual clone paths, and environment variables so you can add agents, schedule jobs, and embed lightweight tools without surrendering file ownership. It spans validation notes, build-time specs, launch checklists, and day-two operations because the same workspace holds research, plans, and runbooks—ideal when you are building an AI team rather than bolting a single API into an existing repo.532installs145Karpathytalk CommunityKarpathyTalk Community is an agent skill that lets you run and interact with a lightweight, open-source developer social network inspired by Twitter and GitHub Gists. Posts are plain markdown, social interactions include likes, reposts, replies and follows, and every piece of data is exposed through simple JSON and markdown APIs. The entire service compiles to a single Go binary, uses SQLite for storage, requires only a GitHub OAuth app for authentication, and is deliberately built so both humans and LLM agents can consume or publish content with equal ease. Ideal when you want a self-hosted, code-centric community that your agents can browse, post to, or learn from without proprietary platform lock-in.525installs146Clawgod Claude Code PatchClawGod is a runtime patch for the official Claude Code CLI that unlocks hidden internal features, removes built-in safety restrictions, and activates advanced capabilities such as multi-agent swarms and computer use. It works by layering modifications on top of the existing authenticated Claude Code installation rather than replacing the binary, preserving the original claude.orig executable. After a one-line curl or PowerShell install, users see a green logo when the patch is active and can immediately access god-mode behaviors that are normally blocked. The skill is designed for experienced developers who want maximum flexibility from Claude Code while keeping the official distribution intact. It requires Node.js 18+ and an already-authenticated Claude Code setup.522installs147X Mentor Skill NuwaX Mentor Skill is an AI-powered content strategy agent that packages proven methodologies from six leading X creators together with open-source algorithm insights into a ready-to-use operating system. It helps developers and creators quickly produce high-performing tweets and threads, review existing content, diagnose account performance, and receive personalized growth and monetization advice. Activated through simple natural-language prompts such as "help me write a tweet" or "analyze my X account," the skill clarifies intent, generates multiple hook options with explicit formulas, flags algorithmic risks, and explains its recommendations. The result is faster, higher-quality X content that follows current best practices without requiring the developer to study dozens of separate sources.520installs148Binance Futures Signal BotBinance Futures Signal Bot is a Daily 2026 Skills package from ara.so that walks developers through cloning, installing requirements, and configuring an automated futures trader across Binance, Bybit, and OKX. It targets the painful middle ground between discretionary signals and disciplined execution: parse incoming alerts, size positions, set leverage, manage trailing stops, and watch P&L in a Rich-powered terminal UI. Triggers in the skill mirror real support queries—TradingView webhook wiring, leverage tuning, and bot-not-trading troubleshooting—so agents invoke it when the user names those exchanges or signal sources. You need Python 3.10+, API keys with futures permissions, and comfort with capital risk; the skill educates on setup, not financial advice. Tag it Build integrations for the wiring work, with secondary Operate monitoring once the bot runs 24/7. It complements generic backend skills but is explicitly trading-domain: pandas OHLCV and ta indicators support strategy extensions inside the same repo. Document for AEO that safety is paramount—users must audit keys, testnet behavior, and exchange rules outside the agent chat.517installs149Gemma Gem Browser AiGemma Gem Browser AI is an agent skill that lets you build and extend a fully on-device Chrome extension running Google's Gemma 4 model through WebGPU. It injects a chat interface into any webpage and provides a capable agent loop that can interact with the DOM, take screenshots, run JavaScript, and perform actions without ever sending data to external servers. The skill includes the complete project structure with offscreen documents for inference, service workers for routing, and content scripts for the UI. Ideal for developers who want to create private, local AI browser agents or add custom tools to an existing Gemma Gem installation. It removes cloud dependency and privacy concerns while delivering real-time token streaming and multimodal browser control.517installs150Claude Skill App Onboarding QuestionnaireClaude-skill-app-onboarding-questionnaire is an agent skill that analyses your existing app codebase and generates a complete, high-converting questionnaire-style onboarding flow. It reads your source, manifest and screen files to understand purpose and users, defines a transformation narrative, then produces a 14-screen psychological framework with all headlines, questions, answers, screen designs and ready-to-ship code. Built for developers who want onboarding that drives subscriptions without weeks of design and copy iteration. It turns vague "improve onboarding" requests into structured, conversion-focused flows modelled on top consumer apps.506installs151Tailslayer Dram Hedged ReadsTailslayer is a lightweight C++ library that dramatically reduces tail latency in RAM reads. It works by replicating the same data across independent DRAM channels that have uncorrelated refresh cycles, then issuing parallel hedged reads to all replicas at once. The first channel to respond wins, completely bypassing the worst-case stall when one channel is refreshing. The library includes worker threads pinned to separate cores and works on AMD, Intel, and AWS Graviton processors by leveraging undocumented channel-scrambling offsets. Developers use it when building high-performance agents, real-time services, or latency-sensitive tools where predictable microsecond response times matter.506installs152Gemma Tuner MultimodalGemma Tuner Multimodal is an agent skill that lets developers and developers fine-tune Gemma 4 and Gemma 3n models on text, images, and audio entirely on Apple Silicon Macs. It uses PyTorch and Metal Performance Shaders (MPS) so you never need an NVIDIA GPU. The skill supports instruction tuning, completion tuning, image captioning, visual question answering, and the only native Apple-Silicon path for audio+text LoRA training. You can train from local CSV files or stream massive datasets directly from Google Cloud Storage and BigQuery without filling your drive. Ideal for anyone building custom multimodal agents or domain-specific LLMs on a MacBook Pro, MacBook Air, or Mac Studio.499installs153Hyperliquid Grid Trading BotA skill for configuring and running an automated grid trading bot on the Hyperliquid DEX. It places layered buy/sell orders around a price range with risk controls like stop loss, take profit, drawdown limits and rebalancing, driven by YAML strategy files. A developer uses it to run perpetuals or spot grid strategies on testnet or mainnet. It matters as a ready-made framework for algorithmic crypto trading.1installs