Skills de Claude Code · página 2

TRL: SFT, DPO, PPO, GRPO, reward modeling for LLM RLHF.

Workout planning, macros, and body metrics via wger/USDA.

Speed up long-sequence transformer training and inference.

Delegate coding to xAI Grok Build CLI (features, PRs).

Constrain LLM output with grammars; guarantee valid JSON.

HeartMuLa: Suno-like song generation from lyrics + tags.

Publish static sites to {slug}.here.now and store private files in cloud Drives for agent-to-agent handoff.

Modify or debug s6 services in the Hermes Docker image.

Configure and troubleshoot Honcho memory for Hermes.

Simplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpeed/FSDP/Megatron/DDP. Automatic device placement, mixed precision (FP16/BF16/FP8). Interactive config, single launch command. HuggingFace ecosystem standard.

HuggingFace hf CLI: search/download/upload models, datasets.

Fast BPE/WordPiece tokenization and custom vocab training.

Render MP4/WebM videos from HTML compositions.

Hyperliquid market data, account history, trade review.

Generate project ideas via creative constraints.

Frontend design guidance, upstream-maintained (impeccable).

Run 150+ AI apps (image, video, LLM) via inference.sh CLI.

Structured LLM outputs validated with Pydantic.

Iterative Python via live Jupyter kernel (hamelnb).

Plan and run multi-agent video production pipelines.

On-demand GPU cloud instances for ML training.

Reserved and on-demand GPU cloud instances for ML training and inference. Use when you need dedicated GPU instances with simple SSH access, persistent filesystems, or high-performance multi-node clusters for large-scale training.

Build leveraged buyout workbooks with IRR/MOIC in Excel.

llama.cpp local GGUF inference + HF Hub model discovery.

Vision-language chat: VQA, captioning, image dialogue.

Manual OAuth for remote MCP servers on headless gateways.

List, auth, and call MCP servers/tools from the terminal.

Create meme PNGs from templates with Pillow text overlay.

>-

Build M&A accretion/dilution workbooks in Excel.

Host modded Minecraft servers (CurseForge, Modrinth).

Serverless GPU cloud for ML jobs and model APIs.

Serverless GPU cloud platform for running ML workloads. Use when you need on-demand GPU access without infrastructure management, deploying ML models as APIs, or running batch jobs with automatic scaling.

Pay HTTP 402 APIs via Machine Payments Protocol (MPP).

Curate LLM training data: dedupe, filter, PII redaction.

Use live BCI cognitive and mood state from NeuroSkill.

OBLITERATUS: abliterate LLM refusals (diff-in-means).

1-3-1 decision briefs: problem, three options, one pick.

Import an OpenClaw setup (memories, skills) into Hermes.

Delegate coding to OpenHands CLI (model-agnostic, LiteLLM).

Optimizes transformer attention with Flash Attention for 2-4x speedup and 10-20x memory reduction. Use when training/running transformers with long sequences (>512 tokens), encountering GPU memory issues with attention, or need faster inference. Supports PyTorch native SDPA, flash-attn library, H100 FP8, and sliding window attention.

Outlines: structured JSON/regex/Pydantic LLM generation.

Fine-tune large LLMs with LoRA on limited GPU memory.

Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70B) with limited GPU memory, when you need to train <1% of parameters with minimal accuracy loss, or for multi-adapter serving. HuggingFace's official library integrated with transformers ecosystem.

Managed vector DB for production RAG and search.

Zero-install localhost tunnels over SSH via Pinggy.

Pixel art w/ era palettes (NES, Game Boy, PICO-8).

Play Pokemon via headless emulator + RAM reads.

Query Polymarket: markets, prices, orderbooks, history.

Build PowerPoint decks headless with python-pptx.

Build creative browser demos with DOM-free text layout.

Expert guidance for Fully Sharded Data Parallel training with PyTorch FSDP - parameter sharding, mixed precision, CPU offloading, FSDP2

High-level PyTorch framework with Trainer class, automatic distributed training (DDP/FSDP/DeepSpeed), callbacks system, and minimal boilerplate. Scales from laptop to supercomputer with same code. Use when you want clean training loops with built-in best practices.

High-performance vector similarity search engine for RAG and semantic search. Use when building production RAG systems requiring fast nearest neighbor search, hybrid search with filtering, or scalable vector storage with Rust-powered performance.

SAM: zero-shot image segmentation via points, boxes, masks.

vLLM: high-throughput LLM serving, OpenAI API, quantization.

Generate a bash wizard guiding a human through manual setup.

Shop catalog search, checkout, order tracking, returns.

Shop.app: product search, order tracking, returns, reorder.

Shopify Admin & Storefront GraphQL APIs via curl. Products, orders, customers, inventory, metafields.

Rewrite text to ASD-STE100 Simplified Technical English.

Simple Preference Optimization for LLM alignment. Reference-free alternative to DPO with better performance (+6.4 points on AlpacaEval 2.0). No reference model needed, more efficient than DPO. Use for preference alignment when want simpler, faster training than DPO/PPO.

SiYuan Note API for searching, reading, creating, and managing blocks and documents in a self-hosted knowledge base via curl.

Throwaway HTML mockups: 2-3 design variants to compare.

Provides guidance for LLM post-training with RL using slime, a Megatron+SGLang framework. Use when training GLM models, implementing custom data generation workflows, or needing tight Megatron-LM integration for RL scaling.

Plan multi-platform social campaigns: briefs to posting.

Query Solana wallets, tokens, txs, and NFTs in USD.

Provides guidance for training and analyzing Sparse Autoencoders (SAEs) using SAELens to decompose neural network activations into interpretable features. Use when discovering interpretable features, analyzing superposition, or studying monosemantic representations in language models.

State-of-the-art text-to-image generation with Stable Diffusion models via HuggingFace Diffusers. Use when generating images from text prompts, performing image-to-image translation, inpainting, or building custom diffusion pipelines.

Stock quotes, history, search, compare, crypto via Yahoo.

Agent payments via Stripe Link — cards, SPT, approvals.

Provision SaaS services + sync creds via Stripe Projects.

Give Hermes phone capabilities without core tool changes. Provision and persist a Twilio number, send and receive SMS/MMS, make direct calls, and place AI-driven outbound calls through Bland.ai or Vapi.

Optimizes LLM inference with NVIDIA TensorRT for maximum throughput and lowest latency. Use for production deployment on NVIDIA GPUs (A100/H100), when you need 10-100x faster inference than PyTorch, or for serving models with quantization (FP8/INT4), in-flight batching, and multi-GPU scaling.

Drive and script tldraw offline canvases with an agent.

Control TouchDesigner via twozero MCP.

Automate Unreal Engine editor scenes, actors, and renders.

Unsloth: 2-5x faster LoRA/QLoRA fine-tuning, less VRAM.

Poll RSS, JSON APIs, and GitHub with watermark dedup.

W&B: log ML experiments, sweeps, model registry, dashboards.

OpenAI's general-purpose speech recognition model. Supports 99 languages, transcription, translation to English, and language identification. Six model sizes from tiny (39M params) to large (1550M params). Use for speech-to-text, podcast transcription, or multilingual audio processing. Best for robust, multilingual ASR.

>

>

>

>

>

>

>

Use only when the user explicitly requests a review or audit of backend code under `api/`. Supports pending-change, file-focused, and pasted-diff reviews. Do not use for implementation-only requests, diagnosis without review intent, frontend code, or backend code outside `api/`.

Refactor high-complexity React components in Dify frontend. Use when `pnpm analyze-component --json` shows complexity > 50 or lineCount > 300, when the user asks for code splitting, hook extraction, or complexity reduction, or when `pnpm analyze-component` warns to refactor before testing; avoid for simple/well-structured components, third-party wrappers, or when the user explicitly wants testing without refactoring.

Use when writing, changing, or reviewing Cucumber and Playwright tests under `e2e/`, including feature files, step definitions, support code, scenario tags, locators, and assertions. Do not use for Vitest, React Testing Library, backend tests, or generic browser automation outside the E2E suite.

Use only when the user explicitly requests a review or audit of frontend code under `web/` or `packages/dify-ui/`. Supports pending-change, file-focused, and pasted-diff reviews. Do not use for implementation-only requests, diagnosis without review intent, or backend-only code.

Use when writing or changing Vitest or React Testing Library tests under `web/` or `packages/dify-ui/`, or when the user explicitly requests frontend test strategy, including evaluation of an existing strategy. Do not use for frontend code-review-only requests, general testability discussion, Python tests, or Cucumber/Playwright E2E.

Use when implementing or refactoring React/TypeScript components and the task requires decisions about component ownership, feature boundaries, state, data flow, effects, or interaction ownership. Do not use for review-only requests, test-only work, copy-only edits, or styling-only changes.

Lightweight coding guardrails for making focused, simple, and verifiable changes in this repo. Use for all coding work.

分享 drawio-chart Skill 的设计思路、安装方式和使用流程，说明为什么技术文章配图更适合保留 draw.io 源文件，以及如何让 Agent 生成可维护的流程图、架构图和模块关系图。

This skill should be used when the user asks to "create an agent", "add an agent", "write a subagent", "agent frontmatter", "when to use description", "agent examples", "agent tools", "agent colors", "autonomous agent", or needs guidance on agent structure, system prompts, triggering conditions, or agent development best practices for Claude Code plugins.

Migrate prompts and code from Claude Sonnet 4.0, Sonnet 4.5, or Opus 4.1 to Opus 4.5. Use when the user wants to update their codebase, prompts, or API calls to use Opus 4.5. Handles model string updates and prompt adjustments for known Opus 4.5 behavioral differences. Does NOT migrate Haiku 4.5.

This skill should be used when the user asks to "create a slash command", "add a command", "write a custom command", "define command arguments", "use command frontmatter", "organize commands", "create command with file references", "interactive command", "use AskUserQuestion in command", or needs guidance on slash command structure, YAML frontmatter fields, dynamic arguments, bash execution in commands, user interaction patterns, or command development best practices for Claude Code.

Guidance for distinctive, intentional visual design when building new UI or reshaping an existing one. Helps with aesthetic direction, typography, and making choices that don't read as templated defaults.