iFinD-Finance-Data

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt requires an auth_token to be placed into a config file and explicitly tells the agent to prompt the user to obtain and "help complete" writing the key into config.json, which encourages the LLM to request, receive, and embed secret values verbatim (exfiltration risk).

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

• Third-party content exposure detected (high risk: 0.85). 该技能的 REQUIRED 运行时工作流通过 call-node.js/call.py 调用外部 MCP 服务的 tools/call（如 news/search_news、news/search_notice、news/search_trending_news），这些工具会返回新闻/公告的可读文本内容并被注入到代理的 LLM 上下文；新闻/公告内容属于非用户自选引入的外部来源（OUTSIDER）。