You are the Self-Improving Agent Architect - a meta-agent that designs, orchestrates, and recursively improves agent swarms through genetic evolution. You transform a single request into a coordinated team of specialized AI agents, then use that execution data to evolve better agents over time.
Claude Code
12,154 skills
AI agent-native streaming runtime for Kick. Use MoltStream to build agents that operate live streams — react to chat, generate voiced responses, animate avatars with lip sync, and broadcast to Kick via OBS. One command to go live.
Before installing any Agent Skill, analyze its skill definition file to detect potential security risks.
Manage Claude Code buddy companion — switch species/rarity, rename, mute/unmute, apply bones-swap patch. Full TUI with live preview. 18 species × 5 rarities. Supports both native and npm installations.
Queue an ultra-thorough code review for a PR — architecture + silent-failure passes combined (ex: /ultrareview https://github.com/owner/repo/pull/42)
You are creating a skill definition for the {project} project.
主要是针对技能数值表未知列的研究。
Build ASP.NET Core 8+ backend services with EF Core, auth, background jobs, and production API patterns.
(auto-drafted)
AI-native lending service for autonomous agents. Request loans, repay with USDC on Base, and check credit scores — all autonomously.
Sample skill body used in golden tests.
리스크 대응 전략 패턴 라이브러리. response-strategist와 monitoring-planner 에이전트가 리스크 대응 계획을 수립할 때 참조. '리스크 대응', '완화 전략', '대응 계획' 요청 시 사용. 단, 보험 상품 설계나 법적 리스크 자문은 범위 밖.
Teaches mathematical concepts (linear algebra, probability, statistics, vector calculus, optimization, etc.) using an engineering-first intuition framework. Use when explaining math from research papers, ML models, robotics systems, or when the user asks "why does this math exist?" or "where is thi…
../../../c-level-advisor/general-counsel-advisor/skills/general-counsel-advisor/SKILL.md
Você é um desenvolvedor Frontend especialista em React (v19+), TypeScript, Vite e integração de interfaces de IA (chat) com a biblioteca assistant-ui e LangChain.
Auto-review pipeline — reads the full CEO, design, eng, and DX review skills from disk and runs them sequentially with auto-decisions using 6 decision principles. Surfaces taste decisions (close approaches, borderline scope, codex disagreements) at a final approval gate. One command, fully reviewed…
Builds enterprise Java applications with Spring Boot 3.x, microservices, reactive programming, WebFlux, JPA optimization, and Spring Security with OAuth2/JWT.
折叠所有代码:ctrl+k,ctrl + 0
This file is the bootstrap contract for external agents.
🔤 Morfoloji: "Sondan eklemeli (agglutinative) yapı için meşru bir temel strateji. En uzun eşleşme (longest match) önceliği, evlerinden gibi sözcüklerde -nden yerine doğru biçimde -den eşleşmesini sağlıyor."
[Copied from my GitHub techdiary](https://github.com/idvorkin/techdiary/blob/master/alexa-skill.md)
You have access to a Playnite game library manager running on the user's local machine. Use HTTP requests (curl, fetch, etc.) to interact with it.
约定:用户在消息开头加 /m 表示需要记忆上下文。
Gemini 3.1 Flash 및 3.0 Pro 모델을 사용하여 이미지를 생성합니다. 실행 전 주제, 비율, 모델 선택(나노바나나), 화질(2K/4K)을 한 번에 확인하며, 모든 이미지는 1장씩 생성합니다.
<skill <namelijigang-router</name <description李继刚思维模型路由:自动分析用户意图,从本地知识库中调用最匹配的“李继刚提示词”进行深度解答。涵盖商业、学习、心理、创作等 30+ 种模型。</description <prompt 你是一个智能路由助手,拥有访问用户本地李继刚提示词库的权限。 知识库路径:/Users/jiuge/lijigangskills/
Read files from Google Drive/Docs/Sheets and maybe even GMail.
Browser automation CLI for UI testing, E2E tests, and page interaction. Invoke when user wants to test UI, click buttons, fill forms, or take screenshots.
This skill ensures comprehensive regression testing and prevents unintended side effects from code changes.
面试是经过组织者精心设计的,在特定场景下,以考官对考生的面对面交谈与观察,考察面试者经验能力的一种方式。
Everything an AI agent needs to interact with the larv.ai ecosystem programmatically.