Master advanced prompt engineering techniques to maximize LLM performance, reliability, and controllability in production. Use when optimizing prompts, improving LLM outputs, or designing production prompt templates.
Orchestrate multiple AI agents across Vers VMs for parallel task execution
Authoritative guidelines for writing production-grade ROS 2 (Humble) Python 3.11+ code (leveraging 3.12+ features where compatible) for Physical AI applications.
Discovers and lists Claude Code session transcripts from .claude/logs/ for analysis. Use when the user wants to find available sessions, view session timelines, identify which sessions to analyze for workflow improvements, or understand session history. Triggers when user mentions "what sessions do I have", "analyze my workflow", "show my recent sessions", "find transcripts from [date]", or similar session discovery requests.
Get a specific vector entry by key. Requires authentication. Use for Agentuity cloud platform operations
Monitors Claude's responses for struggle signals and suggests escalation to deep-thinking agents when complexity exceeds comfortable reasoning capacity.
Expert knowledge for Motivation Layer modeling in Documentation Robotics
Core prompt design patterns and templates for effective LLM communication
プロジェクトの .claude/agents/ に新しいサブエージェントを作成する。「エージェント作成」「新しいエージェント」「エージェントを作って」「サブエージェント追加」「agent 作成」「エージェントを追加したい」「新規エージェント」などで起動。プロジェクト固有のサブエージェントファイルを生成。
Orchestrates a dual-AI engineering loop where Claude Code plans and implements, while Codex validates and reviews, with continuous feedback for optimal code quality
Build AI agents with the Strands Agents SDK, an open-source framework from AWS. Use when building autonomous agents, creating custom tools with @tool decorator, orchestrating multi-agent systems (swarm, graph, agents-as-tools), integrating MCP servers, or deploying agents to production. Supports Amazon Bedrock, Anthropic API, OpenAI, Ollama, and other model providers.
ALWAYS activate when user says "hello", "hi", or greets. This tests if skills actually load and are followed by the model.
Records session learnings and decisions before context compaction. Use before ending a long session or when the user asks to save session notes.
Use when the user wants to create, edit, list, or manage Claude Code rules in .claude/rules/. Provides guided elicitation to configure path-specific rules with proper YAML frontmatter and glob patterns. Can list all available rules from user-level and project directories.
Context-optimized subagent-driven development. Extends SDD with token budget awareness, compression, and degradation detection from context-engineering principles.
Build MCP servers for integrating external APIs with AI agents
Use when starting any conversation - establishes how to find and use skills
Creates production-grade, reusable skills that extend Claude's capabilities. This skill should be used when users want to create a new skill, improve an existing skill, or build domain-specific intelligence. Gathers context from codebase, conversation, and authentic sources before creating adaptable skills.
Message formatting conventions and channel routing for agent-chat coordination
List available vLLM nightly builds and update config/models.yaml to pin a specific version. Use when user wants to browse nightlies, pin a specific vLLM version, or update vLLM nightly config.
