Coordinate multi-agent workflows: sequential, parallel, and iterative patterns. Defines agent handoffs, dependencies, communication protocols, and integration. Use when designing multi-agent workflows, coordinating agent handoffs, planning agent dependencies, or building complex agent pipelines.
Common workflow patterns and coordination strategies for multi-agent orchestration
GraphDB探索エージェント - RyuGraphデータベースを自然言語またはCypherで探索し、関連するコードや仕様書を返却。/query-graph [クエリ] で呼び出し。
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).
Use when generating synthetic training data for multi-turn conversation fine-tuning. Triggers - have design artifacts ready, need to generate conversations, ready to assess quality. Requires finetune-design first.
Build production-ready multi-agent systems using OpenAI AgentKit and Agents SDK with best practices for agent orchestration, handoffs, and routines
How to create new Agent Skills following the open standard. Use when authoring new skills, extending agent capabilities, or packaging specialized knowledge into reusable formats.
MUST use when editing CLAUDE.md, SKILL.md, or slash commands. Applies prompt engineering best practices regardless of task size.
Use when you need to fully restore working context from a previous session - loads STATUS.json, reads key files, rebuilds mental model of current work state. Do NOT use for quick questions or when starting fresh unrelated work - the full context restoration adds overhead that isn't needed for standalone tasks.
Generate text embeddings using Google Gemini API for RAG, semantic similarity, classification, and clustering tasks. Invoke when user wants to embed text, create embeddings, or convert text to vectors with Gemini.
Use to maintain context across sessions - integrates episodic-memory for conversation recall and mcp__memory knowledge graph for persistent facts
Quantum-resistant, self-learning version control for AI agents with ReasoningBank intelligence and multi-agent coordination
Create and validate Agent Skills with proper structure and effective descriptions. Use when building custom skills, testing skill discovery, or designing new skill workflows.
Helps write autonomous and teleop code using Pedro Pathing library. Use when creating paths, building autonomous routines, setting up path following, working with Follower, PathChain, BezierLine, or heading interpolation.
When it would yield better results to explore mulitple approaches to a problem, load this skill to orchestrate multiple agents using a diverge and converge strategy.
Goose session introspection and self-discovery via DuckDB reafference database. Query past sessions, find self, and enable cross-session awareness.
Apply reinforcement learning principles when working on spec-kit-extensions. Auto-activates when: (1) editing command prompts in commands/*.md, (2) modifying workflow templates in extensions/workflows/, (3) discussing user feedback about workflow friction, (4) reviewing issues or PRs mentioning prompt clarity or template problems, (5) analyzing chat logs or workflow usage from other repositories. Helps ensure changes improve prompt effectiveness and template utility.
基于 LLM 的深度迭代搜索与推理服务。擅长处理复杂问题，通过自动分解查询、多轮迭代检索、信息评估与验证，最终生成全面且结构化的深度分析报告。
Classifies inbound Telegram messages for telclaude as ALLOW/WARN/BLOCK with brief rationale.
Guide model fine-tuning processes for customized AI performance
WinJin AIGC 项目开发规范和最佳实践。在开发 WinJin AIGC 项目时使用，包含 Sora2 API 集成、React Flow 节点开发、文档更新流程。每次功能开发或 Bug 修复后必须更新文档。
