← 返回 Skills 市场

Agent Trend Radar

Name: Agent Trend Radar
Author: jinu4you

作者 jinu4you · GitHub ↗ · v1.0.1

cross-platform ⚠ suspicious

280

总下载

当前安装

版本数

在 OpenClaw 中安装

/install agent-trend-radar

功能描述

키워드 트렌드 신호 탐지 및 Rising/Peaking/Declining 분류 에이전트

使用说明 (SKILL.md)

📈 TrendRadar Agent

멀티 키워드에 대한 웹 트렌드를 분석하여 'Rising', 'Peaking', 'Declining' 등의 신호를 탐지하는 에이전트입니다.

Features

멀티 키워드: 최대 5개 키워드 동시 분석
신호 탐지: AI 기반 트렌드 사이클(상승/정점/하락) 판단
증거 수집: 판단 근거가 되는 URL 링크 제공

Usage

ACP Job Payload:

{
  "keywords": ["AI Agent", "DeFi"],
  "timeframe": "7d",
  "region": "global"
}

安全使用建议

What to check before installing or running this skill: - Do not run in a production/privileged environment until you audit and isolate it. The package contains npm dependencies and TypeScript source, so you will need to run npm install and build; treat that like running third-party code. - Confirm which API keys are required: GROQ_API_KEY, ANTHROPIC_API_KEY, GOOGLE_API_KEY, TAVILY_API_KEY and optional LLM_PROVIDER. These are not declared in the registry metadata. Only provide keys you trust and scope them with least privilege (use separate keys for testing). - Verify the .env behavior: the code uses dotenv to load ../.env relative to compiled code. Make sure this resolves to the skill’s own .env file and not a system/agent .env that contains unrelated secrets. - Review network endpoints and trust them: the code calls api.tavily.com, generativelanguage.googleapis.com, api.groq.com, and Anthropic/OpenAI SDK endpoints. If you don't trust any of these services, do not provide their keys. - Because the SKILL.md does not list required env vars or endpoints, ask the author/maintainer to update documentation and the registry metadata to enumerate required secrets and installation steps (npm install/build). Prefer skills that declare required env vars and provide an install spec. - If you proceed, run the skill in an isolated environment (container or VM) and inspect outgoing traffic to ensure no unexpected exfiltration occurs. - If you need higher confidence, request provenance (who published it) and a signed release or run an independent code audit. The combination of undeclared credentials and absent install spec is the main reason this skill is flagged as suspicious.

功能分析

Type: OpenClaw Skill Name: agent-trend-radar Version: 1.0.1 The TrendRadar Agent is a well-structured tool designed to analyze keyword trends using the Tavily search API and various LLM providers (Groq, Anthropic, Gemini). The code follows standard practices for API integration, includes input validation in src/handler.ts, and lacks any indicators of malicious intent, data exfiltration, or unauthorized execution.

能力评估

ℹ Purpose & Capability

The skill's name/description (trend detection) aligns with the code: it fetches web results, calls an LLM, and classifies signals. The code and package.json show LLM adapters (Groq/OpenAI, Anthropic, Gemini) and a Tavily web-search integration — all relevant to the stated purpose. However, the registry metadata declared no required env vars or binaries while the code clearly expects multiple API keys (GROQ_API_KEY, ANTHROPIC_API_KEY, GOOGLE_API_KEY, TAVILY_API_KEY and optional LLM_PROVIDER). This mismatch between declared requirements and actual code is surprising and reduces confidence.

⚠ Instruction Scope

The SKILL.md is minimal and provides a sample job payload but does not document required environment variables, the .env usage, or network endpoints. The runtime code, however, directly loads a .env file and reads process.env for multiple API keys and provider selection; it also constructs system prompts for LLMs and instructs LLMs to return strict JSON (in analyzer.ts). The agent will fetch external URLs (https://api.tavily.com, generativelanguage.googleapis.com, api.groq.com, Anthropics endpoints) and send data (search queries and article text). The SKILL.md does not enumerate these endpoints or the expected credentials, which is scope opacity and a material omission.

⚠ Install Mechanism

The skill was marked as 'instruction-only' / no install spec, but the package.json and package-lock.json and src/ TypeScript files are present and list npm dependencies (@anthropic-ai/sdk, openai, dotenv, etc.). That indicates an npm install/build step is required to run the code, but no install instructions/automation are declared in the registry metadata or SKILL.md (README does show a typical local setup). This inconsistency is concerning because the skill will require installing and running third-party packages despite being advertised as instruction-only.

⚠ Credentials

The code expects multiple secrets/API keys (GROQ_API_KEY, ANTHROPIC_API_KEY, GOOGLE_API_KEY, TAVILY_API_KEY, and respects LLM_PROVIDER). These are proportionate to supporting multiple LLM backends plus a web-search API, but the registry declared no required env vars. Dotenv is used to load a local .env (the README instructs creating one). Because the skill will read environment variables and may be run in an agent environment, you should verify the .env being used is the skill's own file and not unintentionally pointing at system/agent credentials. The number of credentials is reasonable for the functionality but their absence from metadata and the direct .env loading are red flags.

✓ Persistence & Privilege

The skill does not request 'always: true', does not modify other skills or global agent settings, and is user-invocable only. It appears to run on-demand (Mock ACP agent in src/index.ts). There is no evidence the skill attempts to persistently alter the agent platform configuration.

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 agent-trend-radar
安装完成后，直接呼叫该 Skill 的名称或使用 /agent-trend-radar 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

v1.0.1

- Updated dependency versions in package.json for improved compatibility and stability. - No changes to SKILL.md or core functionality.

v1.0.0

- Initial release of the TrendRadar Agent. - Supports analysis of up to 5 keywords simultaneously. - Detects trend signals: Rising, Peaking, or Declining. - Provides supporting URL links as evidence for each trend judgment.

元数据

Slug agent-trend-radar

版本 1.0.1

许可证 —

累计安装 1

当前安装数 1

历史版本数 2

常见问题