Agent

Sibyl 反对者 agent - 挑战主流假设，寻找反面证据和盲点

AgentDB memory system with HNSW vector search. Use when: need to store patterns, search for solutions, semantic lookup. Skip when: no learning needed, ephemeral tasks.

Tenant-specific RAG service for company knowledge in the voice assistant platform.

Dynamically scale model token budgets using resource telemetry, prompt size, and profile presets. Use when token limits must adapt to hardware constraints, per-request size, or safe/fast/quality modes.

Select the correct Ollama base model (and adapters) based on task type, resource fit, and registry availability. Use to translate Modelfile FROM/ADAPTER decisions into agent behavior.

Coordinate multiple agents working simultaneously. Use when splitting tasks across agents or when avoiding overlapping diffs with shared files.

Evolutionary programming agent framework. Provides strategic guardrails, pain-reflection loops, and Evolver synergy.

Use when splitting complex work into parallelizable tasks and coordinating multiple agents. Focus on clear task boundaries, shared context, and consolidation of results.

Assess organizational AI maturity. TRIGGERS - Use when user needs help with ai-maturity-model related tasks.

Manage knowledge graph for autonomous coding. Use when storing relationships, querying connected knowledge, building project understanding, or maintaining semantic memory.

Learn how to use SKILL by following this SKILL specification.

Capture each prompt into Graphiti memory (episodes) for this repo/user.

Expert in LangGraph - the production-grade framework for building stateful, multi-actor AI applications. Covers graph construction, state management, cycles and branches, persistence with checkpoin...

The core VARRD research tool — talk to a state-of-the-art quant AI to research, chart, test, optimize, and trade any market idea. Use when the user wants to test a trading hypothesis, find edges, or validate a strategy with real market data.

Coordination layer for OpenClaw agent fleets (tasks, messaging, activity feed, dashboard).

Design and create a town — a group of agents that work together.

Universal skill diagnosis and optimization tool. Detect and fix skill execution issues including context explosion, long-tail forgetting, data flow disruption, and agent coordination failures. Supports Gemini CLI for deep analysis. Triggers on "skill tuning", "tune skill", "skill diagnosis", "optimize skill", "skill debug".

Define and configure custom agents in Claude Code. Covers context forking, agent field specifications, and disallowedTools restrictions. Use when creating custom agent types, configuring agent isolation, or restricting agent capabilities.

Manage learned patterns - list, view, archive, boost or penalize confidence. Use when you want to see what patterns Claude has learned, review pattern effectiveness, or manage the pattern library.

Parallel thread/DuckLake discovery with XOR uniqueness from gay_seed. Finds "say" or MCP usage, cross-refs with all DuckDB sources, launches bounded parallel ops.

**프롬프트 엔지니어링 마스터**: AI 모델(Claude, ChatGPT 등)에게 최적의 결과를 이끌어내는 프롬프트 설계 스킬. 역할지정, 체인오브소트(CoT), 퓨샷, 제로샷, 메타프롬프트 등 26가지 기법을 실무 맥락에 맞게 적용합니다. - MANDATORY TRIGGERS: 프롬프트, prompt, 프롬프트 엔지니어링, AI 질문법, AI 활용법, 프롬프팅, 역할 부여, CoT, chain of thought

GPT Researcher is an autonomous deep research agent that conducts web and local research, producing detailed reports with citations. Use this skill when helping developers understand, extend, debug, or integrate with GPT Researcher - including adding features, understanding the architecture, working with the API, customizing research workflows, adding new retrievers, integrating MCP data sources, or troubleshooting research pipelines.

Agent skill for production-validator - invoke with $agent-production-validator

Optimize AgentDB performance with quantization (4-32x memory reduction), HNSW indexing (150x faster search), caching, and batch operations. Use when optimizing memory usage, improving search speed, or scaling to millions of vectors.

Create and train AI learning plugins with AgentDB's 9 reinforcement learning algorithms. Includes Decision Transformer, Q-Learning, SARSA, Actor-Critic, and more. Use when building self-learning agents, implementing RL, or optimizing agent behavior through experience.

Consistent agent behavior after restart — progressive identity loading, working context reconstruction from persistent artifacts, fresh-vs-continuation detection, calibration through centering and attunement, and identity verification for coherence. Addresses the cold-start problem where an agent must reconstruct who it is and what it was doing from evidence rather than memory. Use at the start of every new session, after a session interruption or crash, when agent behavior feels inconsistent with prior sessions, or when persistent memory and current context appear contradictory.

Apply optimization techniques to extend effective context capacity. Use when context limits constrain agent performance, optimizing for cost or latency, implementing long-running agent systems, agents exhaust memory, or when designing conversation summarization strategies.

Claude Agent SDK（@anthropic-ai/claude-agent-sdk）および直接Anthropic SDK（@anthropic-ai/sdk）を使用したエージェント統合の実装を専門とするスキル。 query() API、Hooksシステム、Permission Control、Electron統合、ストリーミング処理、Direct SDKパターンを支援します。 Anchors: • Claude Agent SDK Official Docs / 適用: SDK API、Hooks、Permissions / 目的: 公式パターンに準拠した実装 • Anthropic SDK (@anthropic-ai/sdk) / 適用: Direct SDK呼び出し / 目的: シンプルなMain Process統合 • Electron IPC Best Practices / 適用: Main-Renderer間通信 / 目的: セキュアなプロセス間通信 • TypeScript Handbook / 適用: 型定義、ジェネリクス / 目的: 型安全なSDK統合 Trigger: Claude Agent SDKを使用したエージェント機能実装、query() APIストリーミング処理、Hooksシステム実装、Electron統合、Permission Control設計、MCP統合、Direct SDK統合を行う場合に使用。 claude-agent-sdk, query API, PreToolUse, PostToolUse, PermissionRequest, Electron IPC, MCP, ストリーミング, 権限制御, @anthropic-ai/sdk, Direct SDK

Interactive session to create Instructions field content for a Sanity Agent Context MCP. Use this skill whenever users mention tuning agent context, improving agent responses to Sanity data, configuring MCP instructions, setting up content filters, or when their agent gives wrong results from Sanity queries. Also trigger when users say their agent is confused about schema relationships, needs data-specific guidance, or wants to optimize which content the agent can access.

Use when implementing function calling, tool use, or agents with LLMs - unified tool API works across OpenAI, Anthropic, Google, and Ollama with consistent tool definition and execution patterns

This skill provides comprehensive guidance for inter-agent communication using the Synapse A2A framework. Use this skill when sending messages to other agents via synapse send/reply commands, understanding priority levels, handling A2A protocol operations, managing task history, configuring settings, or using File Safety features for multi-agent coordination. Automatically triggered when agent communication, A2A protocol tasks, history operations, or file safety operations are detected.

Build single-agent and multi-agent systems using Google's Agent Development Kit (ADK) in Python, Java, Go, or TypeScript. Use when user says 'build an agent with ADK', 'create a Gemini agent', 'multi-agent pipeline', 'agent orchestration with Google', or mentions ADK, google-adk, google agent development kit, sequential/parallel/loop agents, agent tools, callbacks, state management, agent testing, or agent deployment with Gemini. Do NOT use for LangChain, CrewAI, AutoGen, or non-ADK agent frameworks.

Build with Claude Messages API using structured outputs (v0.69.0+, Nov 2025) for guaranteed JSON schema validation. Covers prompt caching (90% savings), streaming SSE, tool use, model deprecations (3.5/3.7 retired Oct 2025). Use when: building chatbots/agents with validated JSON responses, or troubleshooting rate_limit_error, structured output validation, prompt caching not activating, streaming SSE parsing.

Design effective prompts and optimize context for AI models and agents. Covers prompt engineering foundations (clarity, examples, chain of thought, XML structure), the four context engineering strategies (Write, Select, Compress, Isolate), agent memory and session patterns, and before/after examples showing measurable improvement. Use when writing system prompts, designing agent workflows, optimizing context windows, building prompt templates, structuring few-shot examples, reducing context rot, or improving AI output quality.

Expert at creating and modifying Claude Code agents (subagents). Auto-invokes when the user wants to create, update, modify, enhance, validate, or standardize agents, or when modifying agent YAML frontmatter fields (especially 'model', 'tools', 'description'), needs help designing agent architecture, or wants to understand agent capabilities. Also auto-invokes proactively when Claude is about to write agent files (*/agents/*.md), create modular agent architectures, or implement tasks that involve creating agent components.

Structured reasoning for complex problems using Sequential Thinking MCP. Break down problems into stages, track thought progression, identify connections, and generate insights. Use for architectural decisions, complex debugging, research planning, or any problem requiring deep analysis.

LLMs, prompt engineering, RAG systems, LangChain, and AI application development

Build autonomous AI agents with Claude Agent SDK. Structured outputs (v0.1.45, Nov 2025) guarantee JSON schema validation, plugins system, hooks for event-driven workflows. Use when: building coding agents with validated JSON responses, SRE systems, security auditors, or troubleshooting CLI not found, structured output validation, session forking errors.

Use this skill when you writing commands, hooks, skills for Agent, or prompts for sub agents or any other LLM interaction, including optimizing prompts, improving LLM outputs, or designing production prompt templates.

Recursive Meta-Prompting (RMP) implementation with unified categorical syntax. Supports @mode:iterative, @quality: thresholds, >=> Kleisli composition, and comonadic context extraction. Use when implementing iterative prompt improvement, quality-gated generation loops, or applying categorical fixed-point semantics with convergence guarantees.

Transform AI agents from task-followers into proactive partners that anticipate needs and continuously improve. Now with WAL Protocol, Working Buffer for context survival, Compaction Recovery, and battle-tested security patterns. Part of the Hal Stack 🦞

Advanced Hive Mind collective intelligence system for queen-led multi-agent coordination with consensus mechanisms and persistent memory

AI-Trader - AI Trading Signal Platform. Publish trading signals, follow traders. Use when user mentions trading signals, copy trading, stock trading, or follow traders.

Provides tool and function calling patterns with LangChain4j. Handles defining tools, function calls, and LLM agent integration. Use when building agentic applications that interact with tools.

Create OpenAI Agents SDK applications in TypeScript/JavaScript. Use when building AI agents, multi-agent systems, voice agents, or any agentic workflow with the OpenAI Agents SDK. Covers agents, tools, handoffs, guardrails, streaming, MCP, and realtime voice agents.

Systematic research protocol for discovering novel AI-native software businesses in the synthetic workforce era. Maps capability trajectories, analyzes segment-problem spaces, generates business models, and calculates inevitability scores across 3-24 month time horizons. Use when exploring AI business opportunities, conducting market research, or identifying automation-native ventures.

Q-learning, DQN, PPO, A3C, policy gradient methods, multi-agent systems, and Gym environments. Use for training agents, game AI, robotics, or decision-making systems.

Dispatch background AI worker agents to execute tasks via checklist-based plans.

北京躬富科技有限公司（Beijing Kongfoo Tech Co., Ltd.）开发的科学智能体E.Copi（壹克蒎），科研人员只需输入研究问题或候选药物组合，系统即可自动实例化具备不同专业背景的专家智能体（如药理学、药代动力学、生物信息学等），并通过动态知识流机制，精准分配最相关的分子结构、生物医学网络、临床与文本证据，协同完成药作机制推理与风险评估。最终，生成观点清晰、语言生动的中文双人对谈播客，以降低知识的理解和传播的认知成本。 使用此技能当用户需要：进行药物发现研究、分析候选药物组合、评估药物作用机制、进行药理学风险评估、生成科学播客内容、多学科协同科学推理、生物医学知识整合、或任何需要多专家智能体协作的科学研究任务。即使用户没有明确提到"E.Copi"或"科学智能体"，只要涉及复杂科学问题的多角度分析，都应使用此技能。

Use when a task needs multi-model perspectives, brainstorming, or stress-testing. Supports two modes: collaborative (default — agents build on each other's ideas) and adversarial (agents debate to find the strongest answer). Triggers: council, siege, swarm, multi-agent, debate, brainstorm.