Autonomous multi-session feature development framework. Use when the user wants to systematically implement features from a specification, track progress across sessions, and ensure all tests pass before marking features complete. Activated by the /lynyx-agent-kit:auto-coder command or when working with feature_list.json files in an .auto-coder directory.
Expert in Claude Agent SDK development. Use when users ask about SDK API, agent configuration, MCP servers, hooks, permissions, file checkpointing, or when they mention @AGENT_SDK_DOCS.md. Provides accurate API reference, code examples with TypeScript types, and best practices.
Configure an MCP server for GitHub Copilot or Claude with your Dataverse environment. USE WHEN: "configure MCP", "set up MCP server", "MCP not working", "connect MCP to Dataverse", "add Dataverse to Copilot", "add Dataverse to Claude", "connect via MCP", "use MCP", "MCP tools not available", "no MCP tools", "MCP not configured". DO NOT USE WHEN: workspace not initialized (use dataverse-init first), installing tools (use dataverse-setup).
Set up a machine for Dataverse development — install tools and authenticate. USE WHEN: "install PAC CLI", "install tools", "command not found", "authenticate", "pac auth", "az login", "gh auth", "winget install", "setup machine", "missing tools", "new machine setup". DO NOT USE WHEN: initializing a workspace/repo (use dataverse-init).
ALWAYS LOAD THIS SKILL FIRST for any Dataverse task. Contains hard rules that override all other skills. USE WHEN: ANY request involving Dataverse, Dynamics 365, Power Platform, tables, columns, solutions, records, queries, CRM, metadata, plugins, SDK, Web API, PAC CLI, or environment operations. Also use for: "how do I", "what tool", "which skill", "where do I start", "help with Dataverse", "create table", "create column", "build solution", "query data", "bulk import", "sample data", "support agent", "customer table", "ticket table". This skill MUST be loaded before any other Dataverse skill. Read the Hard Rules section first.
Add a review comment to a GitHub pull request.
This skill should be used when building AI agents using prompt-native architecture where features are defined in prompts, not code. Use it when creating autonomous agents, designing MCP servers, implementing self-modifying systems, or adopting the "trust the agent's intelligence" philosophy.
Full research pass — reads project documents, identifies agents, researches MCS components, designs architecture, enriches brief.json + generates evals. Uses Agent Teams for quality.
Generate an llms.txt file for any project or website following the llmstxt.org specification. Use when asked to create llms.txt, generate LLM-friendly documentation, make a project AI-readable, or prepare documentation for language models.
This skill should be used when the user asks about "Workers API", "fetch handler", "Workers runtime", "request handling", "response handling", "Workers bindings", "environment variables in Workers", "Workers context", or discusses implementing Workers code, routing patterns, or using Cloudflare bindings like KV, D1, R2, Durable Objects in Workers.
Use when generating Java code for web applications, APIs, or enterprise systems - prevents OWASP Top 10 vulnerabilities in Spring Boot, Jakarta EE, and core Java
sdd-initで作成したSpecドラフトを入力として、TDDを回せるだけのテスト観点（ハッピー/代表的失敗/境界/不変条件/非機能）をSpecに網羅する。ここではテスト実装・プロダクション実装は行わない。スコープ変更・要件追加もしない（必要なら未決事項として質問を起票し、合意が取れるまでTDDへ進ませない）。出力は「Specの更新（テスト戦略・テストケース一覧・カバレッジチェック）」「ブロッカー/未決事項」「tdd-redに渡す“次に書くべきテスト1つ”候補」に限定する。
Add error tracking and performance monitoring to your project services. Use this skill when adding error handling, creating new controllers/routes, instrumenting background jobs, or tracking performance. Supports Sentry, Datadog, and other monitoring solutions. ALL ERRORS MUST BE CAPTURED - no exceptions.
Capture and classify learnings from a build/eval/fix session. Compares discoveries against existing knowledge, presents a classified table, and writes approved items to the learnings system.
Report a bug conversationally — Claude gathers details, previews, and creates a GitHub issue via gh CLI.
Create or enhance frontend tasks with detailed implementation steps. Supports TWO modes: MODE A: Enhance existing task (add detailed steps to task in file) MODE B: Create new standalone task file (can contain single or multiple tasks) IMPORTANT: This skill creates MICRO-LEVEL atomic steps, NOT macro architecture plans. For architecture planning or feature scoping, use EnterPlanMode instead. Use this skill when: - Running /frontend.plan command - User asks for "atomic steps", "step-by-step plan with verification" - Have a specific task and need implementation steps - User wants to create a standalone task file (not from story) - Need granular steps: Setup (create files) → Implementation (write code) → Test (verify) DO NOT use for: - Architecture planning (use EnterPlanMode) - General development planning - Feature scoping or estimation Output: - MODE A: Update specific task section in file with Implementation Steps - MODE B: Create new standalone-{name}.md file with one or multiple tasks
Guidelines for maintaining the kanji index feature. Covers kanji ID assignment, index updates, and troubleshooting.
Bundle size reduction strategies, icon optimization, tree-shaking, and performance analysis. Use when reducing bundle size or optimizing imports.
デプロイ済みアプリの作成者レビューを実施する。「レビュー」「改善」「ブラッシュアップ」を依頼されたときに使用する。
Three.js materials - PBR, basic, phong, shader materials, material properties. Use when styling meshes, working with textures, creating custom shaders, or optimizing material performance.