LLM & Agents

This skill should be used when creating custom subagents for Claude Code, configuring specialized AI assistants, or when the user asks about agent creation, agent configuration, or delegating tasks to subagents. Covers both file-based agents and Task tool invocation.

Claude 세션의 대화 요약과 작업 결과를 Notion 페이지 하위에 업로드합니다. '노션 업로드', '결과 저장', 'notion 정리' 요청 시 활성화됩니다.

プロンプトフィードバックレポートを出力する。プロンプトレポートや作業フィードバックとも呼ぶ

Universal workflow coordinator for complex multi-skill and multi-agent operations. Routes requests, manages multi-step workflows, and coordinates sequential processing across Claude skills and GitHub agents.

Create new Claude Code skills with proper structure, validation, and best practices. Generates skills that follow Anthropic specifications and community patterns. Use when you need custom skills for specific workflows, either globally or per-project.

Apply security practices for the .NET 8 WPF widget host app: authN/authZ, secrets, encryption at rest, and secure local storage. Use when handling credentials, tokens, user data, or integrating identity providers.

TOON Format Specialist - Token-efficient data encoding for LLM communication optimized per TOON Spec v2.0

Full-featured Agent Skills management: Search 35+ skills, install locally, star favorites, update from sources. Supports tag search (#azure #bicep), category filtering, and similar skill recommendations.

Execute complex tasks with planning, parallel subagents, atomic commits, and documentation updates. Use when the user asks to implement a feature, refactor code, or complete multi-step work.

Validate Lisp code (Clojure, Racket, Scheme, Common Lisp) for syntax errors, parenthesis balance, and semantic issues. This skill should be used when validating Lisp code files, checking for syntax errors before execution, or validating LLM-generated Lisp code including incomplete or partial expressions. Provides structured JSON output optimized for automated workflows.

Automatically intercepts and optimizes prompts using the prompt-learning MCP server. Learns from performance over time via embedding-indexed history. Uses APE, OPRO, DSPy patterns. Activate on "optimize prompt", "improve this prompt", "prompt engineering", or ANY complex task request. Requires prompt-learning MCP server. NOT for simple questions (just answer them), NOT for direct commands (just execute them), NOT for conversational responses (no optimization needed).

Complete Test-Driven Development workflow for AI-assisted development. Use when writing tests, implementing TDD, or verifying test quality. Includes RED-GREEN-REFACTOR phases, 5 quality criteria, test cheating detection, and coverage requirements.

Perplexity Sonar API development with search-augmented generation, real-time web search, citations, and OpenAI-compatible Chat Completions. Use for AI-powered applications requiring up-to-date information, research assistants, and grounded responses with sources.

Creer un nouveau flow d'automatisation de formulaire complet, incluant exploration du site web, implementation et tests. (project)

Sync ClaudeKit Marketing with Engineer updates using natural language. Handles the two-step sync process, reports what changed, and helps review updates.

CRITICAL - Guide for using Claudish CLI ONLY through sub-agents to run Claude Code with OpenRouter models (Grok, GPT-5, Gemini, MiniMax). NEVER run Claudish directly in main context unless user explicitly requests it. Use when user mentions external AI models, Claudish, OpenRouter, or alternative models. Includes mandatory sub-agent delegation patterns, agent selection guide, file-based instructions, and strict rules to prevent context window pollution.

AGGRESSIVELY use TOON v2.0 format for biggish regular data (≥5 items, ≥60% uniform). Auto-applies to tables, logs, events, transactions, analytics, API responses, database results. Supports 3 array types (inline, tabular, expanded), 3 delimiters (comma, tab, pipe), key folding for nested objects. Triggers on structured data, arrays, repeated patterns. Use TOON by default when tokens matter - RAG pipelines, tool calls, agents, benchmarks. Keywords "data", "array", "list", "table", "log", "transaction", "metric", "analytics", "API", "database", "query", "TOON".

Spawn and manage hierarchical AI sub-agents (Cursor, Codex, Gemini) in isolated git worktrees with explicit role prompts, structured task briefs, and verification checklists. Supports work vs research modes, await vs fire-and-forget runtime, and role-aware wrappers for cleaner delegation. Use when: (1) Delegate tasks to AI agents, (2) Parallel work across features, (3) Research tasks needing web search, (4) Code refactoring or implementation, (5) Documentation updates, (6) Testing agent orchestration systems. Triggers: spawn agent, delegate to agent, agent orchestration, multi-agent, sub-agent, agent work, agent research.

This skill provides comprehensive knowledge for Cloudflare Images, covering both the Images API (upload/storage) and Image Transformations (optimize any image). It should be used when uploading images to Cloudflare, resizing images, optimizing image delivery, implementing direct creator uploads, creating image variants, generating signed URLs for private images, transforming images via Workers, or encountering image upload/transformation errors. Use when: setting up Cloudflare Images storage, implementing user-uploaded images, creating responsive images, optimizing image formats (WebP/AVIF), resizing images via URL or Workers, debugging CORS errors with direct uploads, handling image transformation errors (9401-9413), implementing signed URLs, managing image variants, or building image CDNs. Keywords: cloudflare images, image upload cloudflare, imagedelivery.net, cloudflare image transformations, /cdn-cgi/image/, direct creator upload, image variants, cf.image workers, signed urls images, flexible variants, webp avif conversion, responsive images cloudflare, error 5408, error 9401, error 9403, CORS direct upload, multipart/form-data, image optimization cloudflare

Creates custom output styles for Claude Code that modify system prompts and behavior. Use when the user asks to create output styles, customize Claude's response format, generate output-style files, or mentions output style configuration.