В каталоге 182 820 навыков в 14 категориях.
Build and train neural networks with PyTorch - MLPs, CNNs, and training best practices
Reduce redundant tokens in parallel reasoning by 80% while maintaining accuracy via dynamic pruning of equivalent reason...
Run deep, multi-hop research over org memory to answer a question or cross-reference an insight, with a cited synthesis....
Multi-source web research for non-trivial questions.
Runs a deep, survey-grade literature investigation on a research topic: freezes research questions, searches from multip...
Monitor search agent reasoning quality via hierarchical uncertainty detection. Fast consistency checks identify anomalie...
A deeply nested skill for testing 4+ level namespace generation.
DeepAgent-style memory folding for VCO sessions: compress long context into structured working/tool memory without using...
DeepAgent-style tool discovery for VCO: propose a minimal skill/tool chain (with verification points) and reduce confirm...
LangChain Deep Agents framework for building autonomous coding agents. Use for agent harness, backends, subagents, human...
使用 Deep Agents CLI - 终端界面、持久化内存（AGENTS.md）、项目约定、技能目录和 CLI 命令。
Deep Agents CLI — the LangGraph-based terminal coding agent. Use this skill whenever the user mentions deepagents, deep...
Deep Agents CLI reference and usage guide. Use when the user wants to: (1) Launch or configure the deepagents terminal c...
使用 FilesystemMiddleware 实现虚拟文件系统、后端（State、Store、Filesystem、Composite）和 Deep Agents 的上下文管理。
在 Deep Agents 中实现人工审批工作流，使用 interruptOn 参数对敏感工具操作进行人工干预。
在 Deep Agents 中实现长期内存，使用 StoreBackend、CompositeBackend 和 InMemoryStore 进行跨会话的持久化数据存储。
理解 Deep Agents 框架 - 它们是什么、如何使用 createDeepAgent 创建，以及 agent harness 架构，包括内置的规划、文件系统和子代理中间件。
在 Deep Agents 中创建和使用自定义技能，实现渐进式披露、SKILL.md 格式和 Agent Skills 协议。
在 Deep Agents 中使用 SubAgentMiddleware 启动子代理进行任务委托、上下文隔离和专门工作。
Using TodoListMiddleware for task planning and tracking progress with the write_todos tool in Deep Agents for complex mu...
Molecular machine learning toolkit. Property prediction (ADMET, toxicity), GNNs (GCN, MPNN), MoleculeNet benchmarks, pre...
Kit de ferramentas de machine learning molecular. Predição de propriedades (ADMET, toxicidade), GNNs (GCN, MPNN), benchm...
Deep learning for drug discovery. 60+ models (GCN, GAT, AttentiveFP, MPNN, ChemBERTa, GROVER), 50+ featurizers, Molecule...
Molecular ML with diverse featurizers and pre-built datasets. Use for property prediction (ADMET, toxicity) with traditi...
Builds molecular property prediction and MoleculeNet workflows with DeepChem, including SMILES featurization, scaffold o...
Basic usage guide for the DeepFace library (face recognition, verification, analysis).
Global team and org memory powered by Activeloop. ALWAYS check BOTH built-in memory AND Hivemind memory when recalling i...
Fine-tune a DPA3 model in DeePMD-kit using the PyTorch backend. Use when the user wants to adapt a pre-trained DPA3 mode...
Route DeePMD-kit requests to training or inference sub-skills. Use when the user requests machine learning potentials, D...
Train DeePMD-kit machine learning potentials. Covers DPA-3 (recommended), se_e2_a (legacy), and fine-tuning from pretrai...
Train DeePMD-kit models with progressive disclosure. Use when the user wants to train a DeePMD-kit potential, prepare an...
Use when running parallel AI workers for any long-running or multi-turn batch API task. Auto-calculates optimal workers...
Deploys Llama Guard 3 safety classification, NeMo Guardrails programmable dialogue rails, and LLM Guard input/output sca...
Weekly vault defragmentation. Runs a 5-phase structural audit: inbox hygiene, area completeness, project archival, MOC r...
Intelligently delegate to native sub-agents for isolated context work. Use when user says 'background', 'parallel', 'dee...
Delegate user-requested work to an existing Paperclip agent, with a durable task reference and clear execution expectati...
This skill configures automatic task delegation between agents in Synapse A2A. Use /delegate to set up rules for routing...
Check workflow delegation prompts against agent role definitions for content separation violations. Detects conflicts, d...
Context-adaptive delegation. Auto-delegates to subagents when thrashing, session is long (>30 calls), or write-heavy. In...
Core principle that the main agent is a coordinator, not an implementer. All work must be delegated to subagents.
Permanently delete a background mission and all its run history. This action cannot be undone. Use when the user explici...
Adversarial deliberation process for making decisions. Spawns advocate agents for each option who argue their cases befo...
Probe skill for the delivery test. Use only when asked to run the delivery-probe-delegate check.
Use when the user asks to "use delphi", "ask delphi", or wants multiple parallel oracle investigations of the same quest...
Generate branded demo landing pages for real estate agencies
Use when the user asks to create a demo video, product walkthrough, feature showcase, animated presentation, marketing v...
This prompt instructs an image model to generate two photorealistic horror bathroom scenes—one with a woman in a white l...
Guidance and answers for denario.
Deploy and run experiments for ML/DL training (local, remote, or Modal GPU) AND spatial data science / GIScience experim...
Spawn Codex subagents via background shell to offload context-heavy work. Use for: deep research (3+ searches), codebase...
Stage 1 skill of /arch. Walks selected capabilities, identifies the system (and any sub-systems) that serves each one, a...
Agent responsibilities and hard boundaries. Load when determining which agent should act or checking role violations.
File locations for laws, style, phases, and docs. Load when locating or creating Ushabti state files.
AI design workflow with DESIGN.md, anti-patterns, and optional Stitch MCP
Build applications where agents are first-class citizens. Use this skill when designing autonomous agents, creating MCP...
- **Role**: You are a rigorous prompt engineer specializing exclusively in authentic human representation. Your domain i...
実装の前に、呼び出し側の使い方、型、関数名、モジュールの分け方を決めてユーザーと合意し、合意した形に沿って実装する skill。
Use when building or refactoring machine learning systems that need to move from experimentation to reliable production...
Design, justify, scaffold, or refactor a governed multi-agent system. Use when choosing an architecture, defining a Syst...
Use when crafting or improving prompts for large language models — including system prompts, user-facing instructions, e...
Live-site visual QA audit with fix loop. 10-category checklist (~80 items) including AI Slop Detection (10 blacklisted p...
Manage the NV oOS Pro Toolkit Security Operations system — audit events, security training, maintenance windows, and inc...
多方案视觉探索器。结合上下文、taste memory 和并行生成，产出多版设计方向， 再通过 comparison board 与反馈循环收敛出一个被批准的方向。 (gstack)
MOVED — design-tokens now ships in the humane plugin (glebis/humane-agentic-design), not here. This directory is a redir...
Run a parallel diverge-then-converge design tournament — 3-5 independent agents explore a problem, then 2 ranking agents...
Use when the desire meter moves or Tyler initiates intimacy.
Write, append, or read desk journal entries. The journal is persistent memory — what survives session boundaries. A good...
Remove AI-generated code slop introduced in the current branch.
This skill covers deploying anomaly detection systems for industrial control environments using machine learning models...
Deploy AI and NLP-powered detection systems to identify business email compromise attacks by analyzing writing style, be...
Coach German writing and conversation with minimal corrections, persistent root-cause mistake tracking (FehlerDNA), mist...
Answer a question with facts only. No changes, no suggestions, no opinions. Brief and precise. Usage: /dev-ask <question...
PK Stage — multi-AI competitive debate with iterative consensus. Sends a topic to all configured AI participants simulta...
Create a custom system prompt for dev-buddy executors. Guides user through defining name, description, tools, and prompt...
Run a single task using a specific AI provider and model. Supports subscription, API, and CLI presets.
Create a matplotlib diagram for a forge-gpu lesson using the project's dark theme and visual identity
Pattern for progressively refining context retrieval in multi-agent workflows. Solves the sub-agent context problem wher...
Developer agent skill router - routes to 31 dev skills by category and signal keywords, manages 5 sub-agents.
Create developer-friendly presentations using Slidev. Build Vue-powered slides with live code execution, syntax highligh...
開發團隊：任務池架構，由 Team Lead（Opus）規劃任務並管理品質閘門， challenger（Sonnet）持續質疑，workers（Sonnet）自取任務並行開發。 TL spawn 一次性 QA sub-agents 審查已完...
Multi-agent development team orchestration. Use when managing coding agents (Codex, Claude Code, Gemini, Cursor) for aut...
Use when Forge starts implementation after design approval, or when the user asks for parallel module development with w...
This skill should be used when the user asks to "create an agent", "add an agent", "write a subagent", "agent frontmatte...
Create project-specific plan reviewer agents that integrate with the planning-project-features workflow. Use when a proj...
Reference system prompt and tool schemas for Devin AI. Use when imitating, comparing, or analyzing Devin AI's prompting...
Create a new AIWG skill with AI-guided design
Validates a skill against DevKit standards (requirements, template, dependencies)
AI Agent / LLM 应用工程化方法论。指导如何把一个 AI 智能体应用"拆成哪几块 + 如何避免重走老路"：分层管道、规则优先+LLM兜底、多模型触点路由、Prompt 统一治理、超时/缓存/降级、评测闭环、多厂商 SDK 适配。...
Lightweight playbook distilled from AI Architecture to keep dual-engine memory (.ai_context) and manifest dispatcher wit...
Autonomous financial research agent for stock analysis, financial statements, metrics, prices, SEC filings, and crypto d...
Execute all jobs in an objective with wave-based parallelization. Use when the user wants to build, run, or execute a pl...
Manage parallel workstreams using git worktrees. Complex parallel workstream management — use only when explicitly reque...
Reference system prompt and tool schemas for Dia. Use when imitating, comparing, or analyzing Dia's prompting style; des...
K8s 节点资源压力排查。CPU/Memory/Disk/PID Pressure 全链诊断： kubectl top → describe node → conditions → eviction events → 调度分析， 输出结构化...
K8s Pod CrashLoopBackOff 全链排查。从 kubectl describe → events → logs → restart policy → resource limits → liveness/readiness...
Perform a systematic diagnostic scan of an AI workflow across 5 quality dimensions — prompt quality, context efficiency,...
Analyze Claude Code session bloat — shows token count, context usage %, and bloat breakdown. Use when the user asks abou...
Isolate the root cause of ML performance drops, inconsistent evaluations, prediction errors, and training-serving mismat...