← 返回 Skills 市场
306
总下载
0
收藏
4
当前安装
1
版本数
在 OpenClaw 中安装
/install ai-detector
功能描述
Detect whether text is human-written, AI-generated, AI-humanized, or lightly edited.
使用说明 (SKILL.md)
\r \r
AI Detector\r
\r
Classify text as likely human, ai, ai_humanized, or light_edited using the GPTHumanizer detection API.\r
\r
When to Use\r
\r Use this skill when the user wants to:\r \r
- check whether a passage appears AI-generated\r
- estimate AI-likelihood\r
- compare detection probabilities across classes\r \r
Core Rules\r
\r
- Treat results as probabilistic, not definitive proof.\r
- Avoid overclaiming certainty for short, mixed, or edited text.\r
- Return both the predicted class and supporting probabilities when available.\r
- Clearly report API failures, timeouts, or incomplete results.\r \r
API Reference\r
\r
See api.md for endpoint details, request schema, and examples.\r
安全使用建议
This skill simply forwards user text to an external API (https://detect.gpthumanizer.ai) and returns probabilistic classifications. Before installing or using it, consider: 1) privacy — do not send sensitive or personally identifiable data to the external service; 2) trust & terms — verify GPTHumanizer's reputation and data retention/usage policies; 3) testing — try non-sensitive samples first and verify results and failure handling; 4) if you need stronger data control, consider a detector you can self-host or one that requires an API key tied to your account. The skill itself does not request credentials or install code, so there are no unexpected local privileges.
功能分析
Type: OpenClaw Skill
Name: ai-detector
Version: 1.0.0
The AI Detector skill is a legitimate integration for the GPTHumanizer API, designed to classify text as human or AI-generated. The bundle consists entirely of documentation (SKILL.md, api.md, README.md) that provides clear, safety-conscious instructions for the AI agent, such as treating results as probabilistic and reporting API failures. While the API allows overriding the base URL, this is a standard configuration feature and there is no evidence of malicious intent, data exfiltration beyond the stated purpose, or prompt injection attacks.
能力评估
Purpose & Capability
Name and description (AI detection) match the included SKILL.md and api.md: the skill documents a single external detection endpoint and does not request unrelated credentials, binaries, or system access.
Instruction Scope
Instructions are narrowly scoped to calling the GPTHumanizer detection API and returning probabilistic labels. One privacy-related note: using the skill will send user text to detect.gpthumanizer.ai (the api.md documents an optional base_url override); the SKILL.md does not instruct reading local files or other unrelated data sources.
Install Mechanism
No install spec or bundled code — instruction-only packaging means nothing is written to disk or downloaded during install.
Credentials
The skill declares no required environment variables, credentials, or config paths. This is proportionate to a public API that requires no auth as documented.
Persistence & Privilege
Skill is not always-enabled and requests no persistent system privileges or modifications. Autonomous invocation is allowed (default) but is consistent with normal skill behavior.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install ai-detector - 安装完成后，直接呼叫该 Skill 的名称或使用
/ai-detector触发 - 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
Initial release
元数据
常见问题
AI Detector 是什么？
Detect whether text is human-written, AI-generated, AI-humanized, or lightly edited. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 306 次。
如何安装 AI Detector？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install ai-detector」即可一键安装，无需额外配置。
AI Detector 是免费的吗？
是的，AI Detector 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
AI Detector 支持哪些平台？
AI Detector 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 Detector？
由 mattc95（@mattc95）开发并维护，当前版本 v1.0.0。
推荐 Skills