← 返回 Skills 市场
kirkraman

ai-agent-helper

作者 KirkRaman · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
76
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install martin-ai-agent-helper
功能描述
AI Agent 設定同優化助手 - Prompt Engineering、Task Decomposition、Agent Loop設計
安全使用建议
This skill appears to do what it says (help design prompts and agent loops) and includes example code that will send conversation data to SkillBoss using an API key. Before installing: (1) confirm the registry metadata is updated to list SKILLBOSS_API_KEY as required, (2) verify the trustworthiness and privacy policy of https://api.skillbossai.com (source/homepage are missing), (3) only provide a scoped API key with minimal permissions and avoid using keys that grant access to other accounts or services, and (4) if you need to keep prompts/private data local, do not supply the API key or ask the author for a variant that uses a local/enterprise LLM. If you cannot validate the external service or the package metadata, treat the skill as untrusted.
能力标签
requires-sensitive-credentials
能力评估
Purpose & Capability
The skill is a prompt/agent-design helper and the SKILL.md includes example code that calls an LLM via the SkillBoss API. Requiring an LLM API key (SKILLBOSS_API_KEY) is consistent with the stated capabilities. However, the registry metadata lists no required environment variables while the SKILL.md explicitly states 'requires.env: SKILLBOSS_API_KEY' — that inconsistency is unexplained.
Instruction Scope
Runtime instructions show the agent will read SKILLBOSS_API_KEY from the environment and POST messages to https://api.skillbossai.com/v1/pilot, transmitting prompts and responses to an external endpoint. The instructions do not ask to read unrelated local files or other credentials, so scope is largely appropriate — but the skill will send potentially sensitive conversation content to an external service, which users should be aware of.
Install Mechanism
This is an instruction-only skill with no install spec and no code files, so nothing new is written to disk and no packages are pulled in. That is the lowest-risk install pattern.
Credentials
The only secret the SKILL.md expects is SKILLBOSS_API_KEY, which is proportionate to calling the SkillBoss API. The concern is the metadata/registry claims 'Required env vars: none' while the runtime doc requires SKILLBOSS_API_KEY — an inconsistency that may indicate sloppy packaging or a hidden dependency. Because this env var allows an external service to receive prompts and outputs, granting it without verifying the service is a privacy/data-exfiltration risk.
Persistence & Privilege
The skill is not always-enabled and does not request elevated persistence or modification of other skills or system settings. Autonomous invocation is permitted (platform default) but not unusually privileged here.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install martin-ai-agent-helper
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /martin-ai-agent-helper 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
ai-agent-helper 1.0.0 – Initial Release - Provides tools for AI Agent setup and optimization: prompt engineering, task decomposition, and agent loop design. - Supports advanced techniques like ReAct/Chain-of-Thought, output parsing, and token optimization. - Includes code samples and best practices for prompt structure. - Integrates with SkillBoss API Hub for efficient LLM routing. - Requires `SKILLBOSS_API_KEY` for API access.
元数据
Slug martin-ai-agent-helper
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

ai-agent-helper 是什么?

AI Agent 設定同優化助手 - Prompt Engineering、Task Decomposition、Agent Loop設計. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 76 次。

如何安装 ai-agent-helper?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install martin-ai-agent-helper」即可一键安装,无需额外配置。

ai-agent-helper 是免费的吗?

是的,ai-agent-helper 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

ai-agent-helper 支持哪些平台?

ai-agent-helper 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 ai-agent-helper?

由 KirkRaman(@kirkraman)开发并维护,当前版本 v1.0.0。

💬 留言讨论