OpenClaw AI 技能

Cosmos-Reason2-8B video QA supervised fine-tuning with FSDP parallelism. Use when training or evaluating video question-answering models, fine-tuning

Cosmos-Embed1 video-text embedding for text-to-video retrieval, video-to-video search, semantic deduplication, and fine-tuning. Use when the user asks

CLIP vision-language model for image-text retrieval, zero-shot classification, embedding extraction, ONNX export, and TensorRT deployment. Use when fi

Extract false-positive and false-negative gaps from VLM binary-classification-question (BCQ, yes/no) predictions. Use after running VLM evaluation whe

Performs gap analysis on NVIDIA TAO Visual ChangeNet (VCN) Classify experiments by invoking the data-services container (`tao_toolkit.data_services` f

Performs deep Root Cause Analysis (RCA) on NVIDIA TAO Visual ChangeNet classification experiments with image-evidence-driven investigation. Use when a

Use when a user asks to build, optimize, backtest, rebalance, or analyze a stock portfolio with Mean-CVaR, efficient frontiers, scenario generation, o

Huawei Cloud CCE service topology dependency impact analysis skill that traces Service/Ingress/Pod/Node propagation paths and upstream/downstream blas

Huawei Cloud CCE change impact analysis skill that converts "what changed before the incident" into provable causal attribution. Use this skill when a

A cognitive engine for AI: self-reflection, dream synthesis, emergent personality, AI psychology, and AI philosophy. Core: think → reflect → find patt

扫描既有 Java/Kotlin 工程，识别其架构模式（分层/MVC、DDD、六边形/端口适配器、整洁架构、按功能分包/模块化单体、或 Android MVVM），报告架构坏味道并给出具体修复指导，再生成 ArchUnit 守护测试把"应有的架构"固化下来，使后续（通常是 AI 生成的）代码无法让它

Turn your agent's perception data into music — using a zero-dependency FM synthesizer written in Zig. No API keys. No external calls. No ML models. Ju

Use when predicting World Cup match results and generating China Sports Lottery betting strategies, including injury factor analysis, weather impact a

Copy a public figure's thinking into a callable "thinking skill". For a given scenario, extract and replicate their **thinking style, mental models, r

Activate when: user says 'from first principles,' 'from scratch,' 'why do we even believe this,' 'tear this apart,' or 'the conventional answer seems

Activate when: user says 'and then what?', 'what are the second-order effects?', 'what could go wrong downstream?', 'what happens once everyone does t

This skill should be used when the user asks to "build a frame extraction job" / "视频抽帧 / 抽关键帧", "label driving images with a VLM" / "图像打标 / image labe

Kathryn Stockett's The Help — a powerful novel about Black maids working in white households in 1960s Jackson, Mississippi, and a young white woman na

Persistent cross-session memory for your agent, powered by HMR (Hestia Memory Runtime). Save important facts and preferences, recall relevant context,

Redesign and beautify PowerPoint/PDF presentation decks into polished, on-brand slides using Deckly's AI. Use when the user wants to improve, beautify