industry-stock-tracker

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.75). 脚本在运行时调用远程 API 获取报告 JSON，其中 data.content（以及 wordBase64/pdfBase64 解码后的附件文本）属于外部服务返回的自由文本/文档内容，随后被透传进入 skill 的 LLM 上下文（skill 直接使用脚本 stdout 的 content 字段作为正文）。

HIGH W008: Secret detected in skill content (API keys, tokens, passwords).

• Secret detected (high risk: 1.00). I scanned the script for hardcoded credentials. I found a high-entropy API key string assigned as the default for EM_API_KEY:
• EM_API_KEY default value: "em_fjFqd4YB6Cqs52LF48XWbMDdLNq6MyNg"

This is a literal, non-placeholder value (not like "YOUR_API_KEY" or "sk-xxxx"), looks random/high-entropy, and is used as an authorization header ("em_api_key") for API calls, so it qualifies as a secret.

Other values (API_URL, tool names, output paths, human-readable error messages) are non-sensitive and are not flagged. No PEM/private-key blocks or other high-entropy credentials were present.