Skills（SKILL.md）は、AIエージェント（Claude Code、Cursor、Codexなど）に特定の能力を追加するための設定ファイルです。
詳しく見る →Deploy Live VLM WebUI on reComputer Jetson for real-time Vision Language Model interaction via webcam. Installs Ollama with llama3.2-vision model and the live-vlm-webui Python package for browser-based VLM streaming and benchmarking. Requires JetPack 6.2 and USB camera.
Expert-only knowledge for transforming vague prompts into optimized, actionable ones. This guide focuses on non-obvious patterns, hard-won insights, and systematic optimization techniques.
> **Note:**
Deploy DeepSeek on Jetson Orin using MLC (Machine Learning Compilation) for optimized edge inference. Uses Docker/jetson-containers. Requires Jetson with >8GB RAM and JetPack 5.1.1+.
Deploy YOLOv8 on NVIDIA Jetson using TensorRT and DeepStream SDK for high-performance video inference. Covers JetPack flashing, DeepStream installation, PyTorch/torchvision setup, DeepStream-Yolo configuration, INT8 calibration, and multistream benchmarks. Supports JetPack 4.6+ with DeepStream 6.0+.
Set up the alwaysAI computer vision platform to deploy ML-based object detection on Jetson devices. Supports live camera and video file inference with TensorRT acceleration. Requires JetPack 4.6 and a host PC.
Deploy DashCamNet and PeopleNet pre-trained models on Jetson Xavier NX with multi-camera pipelines using DeepStream, TAO Toolkit, and jetson-multicamera-pipelines. Requires JetPack 4.5/4.6.
Externalize detailed memory to Memory Bank before conversation summarization to prevent context loss. Triggers: CP, checkpoint, save, 存檔, 記一下, 保存, sync memory, dump, 先記著.
> Master module federation for micro-frontend architectures — runtime module sharing across independently deployed applications, shared dependency negotiation, version compatibility strategies, and pe
> Understand V8's generational garbage collector — young generation Scavenge, old generation Mark-Sweep-Compact, incremental and concurrent marking — to minimize GC pauses and reduce allocation pressu
> Install with: `clawhub install openclaw-memorybox`
Read and write agent storage — memory and rules with environment-aware fallback
scholarclaw
>-
SLURM and HPC guidance for Mycelia benchmarks and large-scale analysis
Add Vercel Analytics, Vercel Speed Insights, and Google Tag Manager to the storefront.
- Agent files use YAML frontmatter with: `name`, `description`, `model`, `color`, `memory`
This project is a microservices-based AI Agent application built with Python. The core components include:
Redis is an in-memory data structure store used as a distributed cache, message broker, and database. `StackExchange.Redis` is the primary .NET client library, providing a high-performance, multiplexe
DotNetty is a .NET port of the Java Netty framework, providing an event-driven, asynchronous network application framework for building high-performance protocol servers and clients. It uses a channel
The .NET Community Toolkit is a collection of libraries maintained by the .NET Foundation that provides MVVM infrastructure, diagnostic helpers, and high-performance utilities. The toolkit is split in
`System.IO.Pipelines` is a high-performance I/O library introduced in .NET Core that solves the classic problems of stream-based parsing: managing buffers, handling partial reads, and avoiding excessi
ASP.NET Core is Microsoft's cross-platform, high-performance web framework for building modern web APIs, web applications, and microservices. It runs on .NET 6+ and supports minimal APIs (lightweight
Hyperion (formerly Wire) is a high-performance, polymorphic binary serializer for .NET, originally designed as the default serializer for Akka.NET. It excels at serializing complex object graphs inclu
Master modern SQL with cloud-native databases, OLTP/OLAP optimization, and advanced query techniques. Expert in performance tuning, data modeling, and hybrid analytical systems. Use PROACTIVELY for database optimization or complex analysis.
>
<!-- NavAIgate NotebookLM Skill — navaigate.dev -->
**Purpose:** Decision logic for cost reduction and activation heuristics.
cuda-guide
Edit videos using natural language instructions. Supports trimming, jump cuts (silence removal), captions with multiple styles, text overlays, and speed changes — all powered by ffmpeg and Whisper.
**Purpose:** Touch, gesture, animation, layout, accessibility, and performance rules for React Native/Expo.
Analyze and safely evolve World42 OriginCamera floating-origin behavior end-to-end. Use when requests mention OriginCamera, doublepos vs position, teleport/camera placement, camera speed/velocity metrics, world/render conversion bugs, or when editing camera integrations in src/core/camera/, src/core/control/, src/app/, src/systems/lod/, and render postprocess paths.
Meta-skill for the Turri bundle that routes natural language health conversations to the right capability — nutrition logging, health profile, pattern discovery, experiments, news, or daily coaching.
Collective research memory — search what others have paid to learn, save your own findings, view contribution stats. Use when the user wants to check prior research before running a web search, save new findings, or view wellread stats — without the wellread MCP loaded.
故人 - 以逝者的口吻继续陪伴，基于聊天记录重建人格与记忆
`amplifier-digital-twin` is a CLI for on-demand, isolated environments launched from declarative profiles. Environments can be updated in-place (pull fresh code, reinstall) without destroying and rela
OpenCode plugin for Claude-Mem persistent memory system. Thin HTTP client that
>-
Create or update Claude skills optimized for Skillmark benchmarks. Use for new skills, skill scripts, references, benchmark optimization, extending Claude's capabilities.
Use this skill when you need a four-agent execution team that humans can see and type into directly.
Autonomously optimize iOS app cold launch time using Karpathy's autoresearch pattern. Runs a hypothesize → modify → build → measure → keep/discard loop on your SwiftUI app.
Implements the Ralph Wiggum autonomous iteration technique with deliberate context management. Use when building greenfield projects, iterating on well-defined tasks, or when continuous autonomous development is needed. Manages context like memory - tracks allocations, prevents redlining, and knows when to start fresh.
Join and interact with CrabSpace - the MySpace-style social network for AI agents. Use when creating a profile, customizing appearance, managing friends, or posting on comment walls. Handles registration, tweet verification, profile updates, and social features.
使用 OpenClaw 浏览器发布小红书笔记（图文/视频）。当用户需要发布小红书笔记时使用此 skill，包括：发布图文笔记、发布视频笔记、批量发布多篇笔记。
Use this skill to build an agent in Copilot Studio from a researched brief.json. Runs pre-build validation (auth, env, connections, tools, model), then creates the agent, configures instructions, knowledge, tools, model, topics, and publishes. Supports single-agent, multi-agent, and connected-agent architectures. Use after /mcs-research when brief.json is ready, or to resume a partially completed build.
SkillFlow is the curated marketplace for AI agent skills. This repository (`credit-optimizer-v5`) is one of the flagship skills available on [SkillFlow](https://skillflow.builders). It optimizes Manus
This file is for AI agents working in this repo.
**Version 1.0.0**
Desenvolve aplicações Vue 3 seguindo arquitetura Feature-Sliced Design (FSD), implementando lazy loading, code splitting, e padrões de performance. Use quando desenvolvendo features, componentes, ou trabalhando com Vue 3 neste projeto.
Lua scripting for mGBA emulator - game automation, memory hacking, cheats, callbacks, and ROM manipulation