Skill UI

This guide teaches how to build robust, production-grade ReAct agents using LangGraph 1.0's `create_react_agent`. It specifically addresses common agent failure modes, such as preventing infinite loops (recursion caps) and ensuring proper error propagation. Learn to define typed tools using Pydantic, manage state with checkpointers, and set precise recursion limits. Perfect for users migrating from legacy AgentExecutor or building reliable, multi-step tool-calling workflows.

Master advanced patterns for composing LangGraph subgraphs, which is essential for building robust, multi-agent systems like planner-executors or hierarchical teams. This skill addresses critical issues such as shared state contract propagation (P21) and callback scoping (P28). Learn to manage state across independent graph components using sophisticated dispatch patterns: compiled nodes, Send fan-out, and Command bubble-up.

This powerful CLI allows users to interact with the MiniMax AI platform directly from the terminal. It supports generating and managing diverse multi-modal content, including text, images, video, speech, and music, in addition to performing web searches and analyzing images. It is essential for integrating complex AI capabilities into automated workflows, CI/CD pipelines, or agent-based systems.

This skill facilitates an interactive Quality Assurance (QA) session. Users can conversationally describe encountered bugs or issues. The agent clarifies the problem, explores the codebase for context and domain language, and then meticulously files durable GitHub issues. It supports both single-issue reports and breaking down complex problems into multiple, independently verifiable tickets, ensuring developers receive actionable, high-quality reports written from the user's perspective.

This internal agent skill defines a systematic, multi-step procedure for scanning GitHub repositories to gather comprehensive evidence supporting or explaining bug clusters. The process covers searching open issues, inspecting recent commits, analyzing code paths, and checking deployment history. It assigns evidence confidence tiers (Exact, Strong, Moderate, Weak) to provide a deep, actionable understanding of potential software defects.

Shipwright is an autonomous, multi-phase agent designed to convert simple, plain-English requirements into fully functional, tested, and deployment-ready applications. It supports modern full-stack frameworks like Next.js/Supabase, SvelteKit, and Astro. Use this tool when you need to scaffold a complete project, add complex features, or generate an entire SaaS dashboard merely by describing it.

This tool allows users to execute complex, multi-agent workflows defined in YAML format directly within the AI environment. It functions as an orchestration engine, sequentially invoking various specialized roles (agents) without requiring manual API key configuration. Ideal for structured tasks requiring collaboration between multiple personas.

This skill generates detailed implementation plans for multi-step tasks based on specifications before coding begins. It defines file structures, task granularity, and step-by-step instructions (TDD approach) to ensure clarity and testability. It includes self-check mechanisms and execution handoff options for AI agents or developers.

This workflow establishes a structured, multi-agent system to manage the entire software development lifecycle. It uses specialized roles (Producer, Developer, QA) to clarify scope, implement robust features, conduct rigorous testing, and maintain project context across sessions. Ideal for adopting new projects or coordinating complex, multi-step implementations.

KubeStellar Console is an open-source, multi-cluster Kubernetes dashboard (CNCF project) that provides advanced operational capabilities powered by an MCP server and built-in AI agent skills. It helps manage complex deployments across edge and cloud environments, offering AI-assisted troubleshooting, performance testing, and deep integration with major CNCF projects like Argo and Istio.

This skill generates comprehensive, tailored instruction files for AI coding assistants like GitHub Copilot and Claude. It supports multiple strategies (flat or nested) and is specifically designed for managing complex monorepos by creating area-scoped instructions using `applyTo` globs. Use it to standardize AI tooling usage across large codebases.

This skill implements advanced 'Gatekeeper' logic to optimize AI agent performance in long-running workflows. It prevents context window bloat by sharding metadata, allocating token budgets, and enforcing Atomic Precision to deliver purely functional, filler-free responses. Ideal for complex coding, multi-step reasoning, and large document analysis where conciseness is critical.