memory-rag-instrumentation
Instrument RAG retrieval, memory operations, and context management
smolagents
Build AI agents with Hugging Face's SmolAgents framework. Use when creating code-executing agents, tool-calling agents, multi-agent systems, agentic RAG, text-to-SQL pipelines, web browsing agents, or any multi-step AI workflows. Covers CodeAgent, ToolCallingAgent, custom tools, MCP integration, memory management, secure code execution (E2B, Docker, Blaxel), and model configuration (HF Inference, LiteLLM, Transformers, Ollama).
gemini-image
Generate images from text prompts using fal.ai Gemini 3 Pro. Use when the user asks to create, generate, or make an image from a text description. Supports multiple aspect ratios and resolutions up to 4K.
qino
Ecology for developing ideas through concept exploration, research, and implementation. ACTIVATE for dev work: - "what's next for [app]", "continue [app]", "work on [app]" - "build [feature]", "implement [feature]", "plan the iteration" - "read the implementation notes", "what's the iteration status" - "what should I build next", "next steps for [project]" ACTIVATE for concept work: - "explore [concept]", "go deeper into [idea]", "where am I", "what's here" - "capture this thought", "hold this", "note this" - "test through ecology", "compare these", "attune [quality]" ACTIVATE for research: - "start research on [topic]", "investigate [question]", "begin inquiry" ACTIVATE when in qino workspace (has .claude/qino-config.json). NOT implicit: arc capture requires explicit invocation ("/qino arc" or "capture an arc").
unsloth-dpo
Direct Preference Optimization (DPO) for aligning models with preference data without separate reward models. Triggers: dpo, preference optimization, rlhf, ref_model=none, patchdpotrainer, dpotrainer.
mcp-builder
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
mcp-development
Use when building "MCP server", "Model Context Protocol", creating "Claude tools", "MCP tools", or asking about "FastMCP", "MCP SDK", "tool development for LLMs", "external API integration for Claude"
ai-app
LLM integration patterns for building AI-powered applications with Anthropic Claude and OpenAI APIs. Use when integrating Claude or OpenAI APIs, implementing chat features, adding streaming responses, managing multi-turn conversations, implementing prompt engineering, tracking LLM costs, handling rate limits, or building AI-powered features. Includes client setup, error handling, and cost optimization.
aios-cerebrum
AIOS Agent Operating System and Cerebrum SDK for building, deploying, and orchestrating AI agents
using-superpowers
Use when starting any conversation - establishes mandatory workflows for finding and using skills, including using Skill tool before announcing usage, following brainstorming before coding, and creating TodoWrite todos for checklists
chatbot-dev
ChatBotプロジェクトの開発全般を支援するスキル。プロジェクト構造、コーディング規約、開発ワークフローに関する知識を提供します。ChatBotプロジェクトで作業する時、プロジェクト構造について質問された時、コーディング規約について質問された時、新しい機能を追加する時に使用してください。
stream-chain
Stream-JSON chaining for multi-agent pipelines, data transformation, and sequential workflows
new-proposal
Outline the lightweight process for turning discovery inputs into a first-draft Sentient proposal.
interrogate
Plan and execute an evidence-first interrogation using approved probes; outputs schema-valid findings + report.
model-selection
Automatically applies when choosing LLM models and providers. Ensures proper model comparison, provider selection, cost optimization, fallback patterns, and multi-model strategies.
superthink
Deep reasoning framework for complex tasks with confidence validation, context gathering, and systematic planning
semantic-searchsearch
Use when searching the codebase with natural language queries like "authentication logic" or "database connection"
pr-review
Autonomous multi-agent PR review orchestrator. Spawns parallel workers to process all AI review comments until PR is ready for merge. Supports CodeRabbit, Gemini, Codex, Sourcery, Qodo, Copilot agents. Triggers: "review PR", "PR review", "process comments", "AI review", "orchestrate review", "review cycle"
run-synth-server
Instructions for starting the MCP Synth server (FastAPI) for synthetic data generation and tools.
