LLM & Agents

Manage Claude Code sessions for the current branch. Use this skill when: - User asks "what's next?" or "where was I?" - User starts work on a new branch without a session - Before committing (to update session log) - When resuming work after a break

Strategic planning and task orchestration agent for breaking down complex tasks into actionable execution plans

AI agent development with LangChain, CrewAI, AutoGen, and tool integration patterns.

Deploy GitHub repositories to remote servers via SSH. Claude reads the README, understands build/start instructions, and executes deployment commands over SSH. Supports rollback on failure.

Multi-source comprehensive research using claude-researcher and gemini-researcher agents. Three modes - Quick (3 agents), Standard (9 agents), Extensive (24 agents with be-creative skill). USE WHEN user says 'do research', 'quick research', 'extensive research', 'find information about', 'investigate', 'analyze trends', 'current events', or any research-related request.

AI swarm orchestration and management in Flow Nexus cloud. Use for deploying, coordinating, and scaling multi-agent swarms for complex task execution.

Build mithril-core shared infrastructure. Use when implementing storage, compression, hashing, or common types for the Mithril ML toolkit.

Unit testing patterns with xUnit, Moq, and FluentAssertions including AAA pattern, naming conventions, mocking, and edge case coverage. Use when writing or reviewing unit tests.

This skill should be used when users want professional SDLC practices with automated git workflows, intelligent code review, and project management without rigid processes. Orchestrates specialized agents for feature development, branch management, language-specific code review, and GitHub integration. Ideal for developers who want to stay in flow while maintaining code quality.

복잡한 판단이 필요할 때 decision-agent 활용 가이드

Analyzes warehouse shipment backlogs to optimize batch picking and packing workflows. Use when a merchant needs help clearing order backlogs, improving picking efficiency, optimizing batch sizes, or understanding order composition patterns. Triggers on questions about shipping delays, picking strategies, packing station setup, warehouse workflow optimization, or order fulfillment bottlenecks. Works with Fulfil MCP data or user-provided shipment data (CSV/Excel).

Extend the Strategy Coach agent with new capabilities. Use when user says "add coaching phase", "new research pillar", "extend agent", "add strategic output", or asks to enhance the AI coaching methodology.

AI agent with retrieval tool for document Q&A using RAG and LangGraph.

Use when receiving code review feedback (especially if unclear or technically questionable), when completing tasks or major features requiring review before proceeding, or before making any completion/success claims. Covers three practices - receiving feedback with technical rigor over performative agreement, requesting reviews via code-reviewer subagent, and verification gates requiring evidence before any status claims. Essential for subagent-driven development, pull requests, and preventing false completion claims.

Implement real-time streaming UI patterns for AI chat applications. Use when adding response lifecycle handlers, progress indicators, client effects, or thread state synchronization. Covers onResponseStart/End, onEffect, ProgressUpdateEvent, and client tools. NOT when building basic chat without real-time feedback.

Intent-based code discovery for CLI AI agents using semantic search MCP tools. Use when finding code by what it does (not what it's called), exploring unfamiliar areas, or understanding feature implementations. Mandatory for code discovery tasks when you have MCP access.

Expert guidance for implementing BRC-100 conforming wallets using @bsv/wallet-toolbox. Covers wallet initialization, transaction creation/signing, key management, storage, and certificate operations following the BRC-100 standard.

Core fundamentals of Unsloth for fast LLM fine-tuning, covering FastLanguageModel setup, optimized gradient checkpointing, and native inference acceleration (triggers: unsloth, FastLanguageModel, from_pretrained, get_peft_model, for_inference, gradient checkpointing).

Use when user says "execute plan", "implement story", "run plan for [ID]", "start implementation", or asks to execute a planned story - loads TDD implementation plan from .claude/data/plans/, executes RED-GREEN-COMMIT cycles for each task, updates story status through active->reviewing->implemented, verifies acceptance criteria, and outputs implementation report. (project)

Comprehensive AI/ML development guide for LangChain, LangGraph, and ML model integration in FastAPI. Use when building LLM applications, agents, RAG systems, sentiment analysis, aspect-based analysis, chain orchestration, prompt engineering, vector stores, embeddings, or integrating ML models with FastAPI endpoints. Covers LangChain patterns, LangGraph state machines, model deployment, API integration, streaming, error handling, and best practices.