kadaliao

Claude Code Skills Collection — Development skill for Claude Code

Development community

8 Claude Code Agent Skills from ChinaSiro/claude-code-sourcemap — adversarial-verification, smart-memory-guard, lightweight-explorer, coordinator-orchestrator, worker-prompt-craft, memory-type-system.

How to install Claude Code Skills Collection

This entry records only its repository, not the path inside it, so there is no exact command to give. Open kadaliao/claude-code-skills-collection and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Claude Code Skills Collection does

Alternatives in Development

05 Coordinator System Prompt — Coordinator System Prompt 2.3k ★
SubAgent — You are sub-agent {index} of {total} inside Joanium - a focused research worker 250 ★
Claude Code Sourcemap — This is a mirror file of the leaked source code of Claude code（2026.3.31） 206 ★

README

Claude Code Skills Collection

从 [ChinaSiro/claude-code-sourcemap](https://github.com/ChinaSiro/claude-code-sourcemap) 泄露仓库收束整理的 8 个 Agent Skill，覆盖 Claude Code / AI Agent 的核心工程模式。

Skills 列表

Skill	描述
adversarial-verification	🛡️ 对抗性验证 — 不是确认"它能工作"，而是尝试"把它搞坏"
smart-memory-guard	🧠 记忆防膨胀 + 防漂移 — 解决 AI Agent 记忆文件无限增长、过时记忆误导行动的问题
lightweight-explorer	🔍 轻量探索模式 — 用最少 token 完成代码搜索和调研
coordinator-orchestrator	🎯 多 Agent 协调者模式 — 调度 Worker 做事，负责综合和决策
worker-prompt-craft	✍️ 子代理 Prompt 编写指南 — 写出让子代理一次做对的 prompt
memory-type-system	📝 结构化记忆类型系统 — 四类型分类 + frontmatter 规范 + 写入/召回/漂移防护全流程
self-rationalization-guard	🪞 自我合理化防护 — 识别和反制 AI Agent 常见的偷懒、逃避、自我欺骗模式
task-concurrency-patterns	⚡ 任务并发与依赖管理 — 多 Agent/多步骤任务的并发安全、读写分离、依赖编排