AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Agent Creator

skill-caomeiyouren-cmyr-skills-agents-agent-creator · by CaoMeiYouRen

创建、重构、审计和治理本项目自定义 agent 时使用。用户提到 create agent、new agent、build agent、write .agent.md、custom agent、agent mode、agent workflow、agent governance、判断应该做成 skill 还是 agent、审计 agent 安全与准入标准时都应触发。它负责先判断该需求应由 skill 还是 agent 承载,再在需要时创建或更新 agent,并审计重叠、规范与安全。

No reviews yet
0 installs
34 views
0.0% view→install

Install

$ agentstack add skill-caomeiyouren-cmyr-skills-agents-agent-creator

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-caomeiyouren-cmyr-skills-agents-agent-creator)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
2mo ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Agent Creator? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Agent Creator

铁律:不要先写新 agent,再思考它是不是其实应该是 skill,或者只是现有角色的一次补充。

工作流

  • [ ] Step 1: 判断应由 skill 还是 agent 承载 ⚠️ REQUIRED
  • [ ] 1.1 加载 references/admission-guide.md,先判断这是可复用工作流,还是需要上下文隔离与多角色编排。
  • [ ] 1.2 如果 skill 更合适,停止新建 agent,转交 skill-creator。
  • [ ] Step 2: 审计现有能力 ⚠️ REQUIRED
  • [ ] 2.1 扫描 AGENTS.md、agents/ 与 skills/,确认是否已有高重叠角色。
  • [ ] 2.2 如果只是现有 agent 或 skill 的补充,优先补到原文件,不新增目录。
  • [ ] Step 3: 设计 agent 边界
  • [ ] 3.1 明确 agent 的职责、输入、输出、接棒关系和不负责的事项。
  • [ ] 3.2 明确它集成哪些 skills,是否需要调用其他 agents。
  • [ ] Step 4: 编写或更新 .agent.md ⚠️ REQUIRED
  • [ ] 4.1 保持 frontmatter 可触发,正文包含角色设定、集成技能、参考文档、核心职责和协作工作流。
  • [ ] 4.2 不使用不存在的文档路径作为强制参考资料。
  • [ ] Step 5: 审计安全与准入
  • [ ] 5.1 检查是否与现有 agents / skills 高度重复。
  • [ ] 5.2 检查是否引导越权操作、危险命令或模糊职责。
  • [ ] 5.3 需要审查时,联动 code-reviewer 做结构与安全审计。
  • [ ] Step 6: 同步项目索引
  • [ ] 6.1 更新 AGENTS.md 中的 agent/skill 列表与定位说明。
  • [ ] 6.2 如果新增的是治理型角色,明确其准入规则与管理边界。

核心规则

  • 需要上下文隔离、多角色协作、分阶段接棒或专门治理职责时,优先考虑 agent。
  • 只是一个可复用工作流、检查清单或带 references/scripts 的任务时,优先考虑 skill。
  • 发现与现有能力高重叠时,默认补充现有 skill 或 agent,而不是创建新项。

反模式

  • 用 agent 解决其实是 skill 更适合的问题。
  • 为了“看起来完整”创建一个和现有角色几乎重复的新 agent。
  • 把 agent 写成一个没有边界的万能执行器。
  • 在 agent 里引用不存在的 docs/ 路径或虚构流程。

交付前检查

  • [ ] 已明确这是 skill 问题还是 agent 问题。
  • [ ] 已审计现有 skills / agents,确认不是高重叠新增。
  • [ ] .agent.md 中职责、边界、接棒关系和集成技能清晰。
  • [ ] 已同步 AGENTS.md 或其他必要索引。

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.