数据分析与反馈

数据分析与反馈技能 - 投放效果数据采集、分析与模型迭代建议

Security Scan

The skill's description, README, and most of the TypeScript implementation are coherent for an analytics/reporting tool, but part of index.ts is truncated in the package you provided so I cannot fully confirm there is no hidden network, credential, or file-access behavior.

Purpose & Capability

Name/description (ad analytics, feedback) match the provided code and SKILL.md. The code implements ingest, reporting, feedback generation and campaign management; none of the declared requirements (no env vars, no binaries, no install) are surprising for this purpose.

Instruction Scope

SKILL.md contains concrete action invocations (analytics-feedback.collect/analyze/report) and does not instruct the agent to read arbitrary system files or unrelated environment variables. However, part of index.ts was truncated in the supplied text (the checkPerformanceAlerts and subsequent logic are cut off), so it is not possible to confirm that runtime behavior stays fully within the documented scope (e.g., there might be hidden telemetry or external calls in the truncated portion).

Install Mechanism

No install spec (instruction-only + TypeScript source). package.json only lists build/dev deps. Nothing is downloaded from external URLs and no install-time arbitrary code execution is declared.

Credentials

The skill declares no required environment variables, primary credential, or config paths, and the visible code does not reference process.env or other credentials. This is proportionate to an analytics/reporting skill as presented.

Persistence & Privilege

always is false and there is no indication the skill requests persistent system-wide privileges or modifies other skills. The skill stores its own in-memory state (campaigns, performance, reports) which is expected for this functionality.

What to consider before installing

This package looks coherent for ad analytics and reporting, but you should not install it as-is without a full source review because the provided index.ts was truncated. Before installing or enabling the skill: (1) obtain and review the complete index.ts to ensure there are no network calls, secrets access, telemetry endpoints, or filesystem access in the truncated code; (2) run the skill in a sandbox or dev environment first and monitor outbound network traffic; (3) confirm what data sources it will ingest and that you are comfortable sending those datasets to this code (especially PII or commercial data); (4) if you need stronger guarantees, require the publisher to provide a signed release or host the package on a trusted registry; (5) remember the skill can be invoked autonomously by the agent by default — only enable that if you trust its behavior. If you can provide the full, untruncated index.ts I can re-evaluate and likely raise confidence.

Analytics Feedback - 数据分析与反馈

负责广告投放效果的数据采集、分析和反馈，为模型优化提供建议。

Setup

无需额外依赖，TypeScript编译后使用。

When to Use

• 收集广告投放数据
• 分析投放效果
• 生成数据报告
• 提供优化建议

Architecture

analytics-feedback/
├── index.ts          # 主入口，数据处理和分析
├── package.json      # 依赖配置
└── README.md         # 详细文档

Core Commands

采集数据

await api.executeAction('analytics-feedback.collect', {
  campaignId: string,   // 广告活动ID
  startTime: string,    // 开始时间
  endTime: string,      // 结束时间
  metrics?: string[]    // 指标列表
});

分析数据

await api.executeAction('analytics-feedback.analyze', {
  data: any[],          // 原始数据
  analysisType?: string // 分析类型
});

生成报告

await api.executeAction('analytics-feedback.report', {
  analysisId: string,   // 分析ID
  format?: string      // 报告格式
});

响应事件

• analytics-feedback.data-collected - 数据采集完成
• analytics-feedback.analysis-complete - 分析完成
• analytics-feedback.report-ready - 报告就绪