В каталоге 182 820 навыков в 14 категориях.
Generate YouTube thumbnail prompts that transform freeze frames into cinematic thumbnails.
Generate article or newsletter thumbnail candidates using the Gemini API from inside Claude Code. Claude reads article c...
Generate thumbnails from images with smart cropping, multiple sizes, and batch processing. Ideal for web galleries, soci...
YouTube thumbnail generator via Higgsfield Nano Banana Pro. Claude reads the video's hook/title (Notion page or raw stri...
Generate prompts for dev.to blog thumbnail/cover images in hand-drawn infographic style. Use when creating cover images,...
天蚕土豆网文创作思维——退婚流创始人、升级进度条可视化、换地图循环、废柴逆袭黄金三角、"恐怖如斯"战斗公式、六波推动IP。 触发词：「天蚕土豆视角」「像天蚕土豆那样写」「斗破苍穹」「退婚流」「升级流」「恐怖如斯」。 擅长：退婚流创作、升级流...
天下霸唱盗墓开山鼻祖思维——鬼吹灯、说书人风格、风水秘术、民俗怪谈。 触发词：「天下霸唱视角」「像天下霸唱那样写」「鬼吹灯」「摸金校尉」「寻龙诀」。 擅长：盗墓探险、风水秘术、民俗怪谈、说书人叙事、章回体结构。
天涯神贴大神智囊团 — 20位传奇大神组成的多Agent决策系统。 基于ClawTeam多Agent协作原理，实现「一个目标 → 20位大神并行分析 → 综合决策」的完整工作流。 核心机制： - 智能领导者Agent（协调者）接收目标，分...
R machine learning with tidymodels: recipes (preprocessing), parsnip (model specs), workflows (fit pipelines), tune (hyp...
The cost-discipline constitution for multi-agent work: the **session's top model orchestrates; cheaper tiers execute**.
---name: time-resolved-cryoem-agent
Análisis de series temporales — tendencia, estacionalidad, forecasting con Prophet, statsmodels, y Machine Learning. Tri...
Time series forecasting with ARIMA, Prophet, LSTM, and statistical methods. Activates for "time series", "forecasting",...
Creates an SVG time card showing the current time for Dubai. Writes the SVG to agent-teams/output/dubai-time.svg and upd...
Run an authorized task inside an Available Work Time (AWT) window with a shorter Closeout Grace Period (CGP), fixed dead...
Generate a chronological timeline for a project or topic using sub-agents.
Check whether an opportunity's LLOs are hitting milestones on schedule and email a nudge to any that are behind.
Analyze time-series RNA-seq data using limma voom with splines, maSigPro, and ImpulseDE2. Identify genes with dynamic ex...
Zero-shot time series forecasting with Google's TimesFM foundation model. Use this skill when forecasting ANY univariate...
Zero-shot time series forecasting with Google's TimesFM foundation model. Use for any univariate time series (sales, sen...
Cambridge Quantum (Quantinuum) t|ket> compiler skill for platform-independent circuit optimization
Tool integration agent loops iterative repair and advanced workflows for TLA+ including SANY TLC TLAPS grammar constrain...
Example Marathon Configuration for CLAUDE.md - copy the section below into your project's CLAUDE.md
---name: tme-immune-profiling-agent
Miyabi tmux マルチエージェント・オーケストレーションとiTerm2プロファイル統合。Agent別カラーテーマ、状態監視、自動プロファイル切り替えを提供。
PM-posture terminal pipeline step. Validate the developer marker and real deferrals, take main's advance and run the sui...
Tod des Mieters im Wohnraummietrecht – Eintrittsrecht von Ehegatten, Lebenspartnern, Kindern, Familienangehörigen und Ha...
Turn the recommendations or action list from the preceding assistant response into a numbered todo list in the recommend...
Sweep the user's todo file (TODO.<git user.name>.md, resolved as in the todo rules) at the end of a session: replace the...
Cloud platform for running open-source AI models. Provides inference APIs for LLMs, image models, and embedding models....
Active ou desactive la session memory sur le projet courant. Si active, desactive et supprime le fichier. Si inactive, a...
Proactive token budget assessment and task chunking strategy. Use this skill when queries involve multiple large file up...
Analyze static code for token usage patterns, waste, and budget risks. Trigger: /token-budget
Use when the user asks about token usage, session cost, or wants to check optimization effectiveness.
- Run at the start of every session and whenever prompt sizes or tool calls begin to spike.
Milo's in-session operating discipline for maximizing useful work per token and preventing context/credit drain — the co...
Compressed communication using symbols and abbreviations. Use when context is limited or brevity is needed.
Reduce token waste by 40-60% through anti-sycophancy rules, tool-call budgets, one-pass coding, task profiles, and read-...
Auto-injects token-efficient delegation patterns. Claude tokens are 50x more expensive than MiniMax/GLM/Kimi.
Convert verbose docs/markdown/text into token-efficient formats. Use when user wants to reduce token count, compress con...
尚书省首次调用时加载的 Token 优化策略，定义运行模式拆分、会话复用与上下文管理
CRITICAL - Read FIRST before any work. Strategies to minimize token usage and reduce costs by 60-80%.
Reduce token count in prompts, docs, and prose. Covers prompt compression (40-60% savings), doc formatting, TOON data se...
Port of .claude/commands/token-plan.md for Codex.
Token usage monitoring and cost tracking with configurable thresholds. Reference this skill to understand pricing and li...
When user says "@token-usage", "token usage", "context window", or "token cost".
Build and share a visual dashboard for OpenClaw token usage, model cost comparison, and recent model change timeline. Us...
BPE, WordPiece, SentencePiece, and Unigram tokenizer training, vocabulary optimization, domain extension, and multilingu...
Write or fix prompts for the model families in the TokenLab catalog: FLUX, Nano Banana / Gemini image, GPT image, Seedre...
Check tool/function definitions for hallucination-prone schemas. Trigger: /tool-fidelity
Generate and edit AI images through the local ComfyUI API (FLUX) — txt2img, img2img variation, 4x upscale, AnimateDiff i...
Build tools that agents can use effectively, including architectural reduction patterns
This skill should be used when the user asks to "design agent tools", "create tool descriptions", "reduce tool complexit...
Build tools that agents can use effectively, including architectural reduction patterns. Use when creating new tools for...
Design tools that agents can use effectively, including when to reduce tool complexity. Use when creating, optimizing, o...
Guide for discovering and using MCP tools via the discover_tools meta-tool. Use this skill when asked to perform tasks t...
Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is th...
CLI-first tool selection policy for Claude Code. Use when choosing between CLI tools and MCP servers, designing new agen...
Design systematic decision frameworks for selecting appropriate AI tools (Claude Code vs Gemini CLI) based on context re...
当需要为 AI 定义工具接口、设计调用规范、实现工具发现与编排机制时调用此 skill。典型场景包括：设计 AI agent 的工具集、定义 JSON Schema/XML/TypeScript 格式的工具描述、实现工具权限控制与并行调度、...
Use skills for domain knowledge, MCP tools over manual lookups
Tool use patterns for Claude including schema design, tool_choice modes, result handling, parallel execution, error reco...
Function definition schemas, API sanitization, and execution constraints.
Patterns for Claude tool use including tool definition schemas, multi-tool orchestration, parallel tool calls, error han...
pip install anthropic
Writes and fixes Vercel AI SDK (npm package `ai`, version 7) code for tool calling, multi-step tool loops, agents and st...
Capture session learnings and propose targeted improvements to docs, skills, or core agent instructions. Evidence-based...
Use this skill when working with scientific research tools and workflows across bioinformatics, cheminformatics, genomic...
Build AI scientist systems with the ToolUniverse Python SDK for scientific research. Covers the 3 calling patterns (`tu....
Token-Oriented Object Notation (TOON) format for LLM-optimized data encoding. Converts JSON to compact, human-readable f...
TOON (Token-Oriented Object Notation) encoding for LLM-efficient data representation. 30-60% token savings vs JSON for s...
Expert skill for Token-Oriented Object Notation (TOON) — compact, schema-aware JSON encoding for LLM prompts that reduce...
Token-Oriented Object Notation (TOON) format expert for 30-60% token savings on structured data. Auto-applies to arrays...
Formats structured data using TOON v2.0 to minimize tokens while preserving readability. Use when outputs include tables...
TOON format knowledge and usage patterns for agent communication and memory persistence in plan-marshall marketplace
Identifies and visualizes the top expressing genes per cluster in T/B cells, followed by pathway enrichment analysis. Pr...
Identifies and visualizes the top expressing genes per cluster across ALL cells (before T/B cell selection), followed by...
Research State of the Art (SOTA) literature for an Artificial Intelligence / Machine Learning (AI/ML) topic, method, or...
Extract topics from text collections using LDA (Latent Dirichlet Allocation) with keyword extraction and topic visualiza...
Separates T and non-T cells or B and non-B cells from a mixed cell population. Uses either clonotype percentage from VDJ...
Use torch.compile to JIT-compile PyTorch code into optimized kernels, then validate speedups with warmups and graph-brea...
PyTorch Geometric (PyG) for graph neural networks — node/link/graph classification, message passing (GCN, GAT, GraphSAGE...
Audio signal processing library for PyTorch. Covers feature extraction (spectrograms, mel-scale), waveform manipulation,...
Graph-based drug discovery toolkit. Molecular property prediction (ADMET), protein modeling, knowledge graph reasoning,...
PyTorch-native graph neural networks for molecules and proteins. Use when building custom GNN architectures for drug dis...
Provides guidance for PyTorch-native agentic RL using torchforge, Meta's library separating infra from algorithms. Use w...
Export torchscript exporter operations. Auto-activating skill for ML Deployment. Triggers on: torchscript exporter, torc...
Natural Language Processing utilities for PyTorch (Legacy). Includes tokenizers, vocabulary building, and DataPipe-based...
Provides PyTorch-native distributed LLM pretraining using torchtitan with 4D parallelism (FSDP2, TP, PP, CP). Use when p...
Decode a terms of service or privacy policy into what you're actually agreeing to, ranked by real-world impact. Use when...
The only memory skill that watches on its own. No database. No vectors. No manual saves. Just an LLM observer that compr...
Create or update agent skills. Use when creating new skills from scratch, refactoring existing skills, or validating ski...
Friendly onboarding tour of Claude Code capabilities for users asking what it can do.
Use when the user wants an autonomous ML research loop that pressure-tests competing ideas before spending compute — sev...
---name: tpd-ternary-complex-agent
显示 Agent 流程追踪时间线与摘要
Automatically trace Claude Code conversations to Braintrust for observability. Captures sessions, conversation turns, an...
Use this to add observability to a multi-agent or agentic system (multiple agents, sub-agents, tool loops, handoffs). Tr...
Investigate where a Hermes skill came from, when it was created, and whether it was builtin, promoted/imported, or local...
Imported skill tracing from langchain