wecom-smartsheet

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). It instructs executing JS in a logged-in page and returning raw cell values via JSON.stringify(rows), which could include API keys or passwords verbatim and thus leads to secret exfiltration risk.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This content explicitly describes and documents a method to bypass official API access controls by reading in-browser memory models to extract arbitrary enterprise WeCom (智能表格) data (including cross-sheet records and user info), which enables unauthorized access and straightforward data exfiltration.

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

• Third-party content exposure detected (high risk: 0.80). The skill's SKILL.md explicitly instructs using the browser tool to execute window.getPreloadedTablesManager() and read table.views.recordIds and table.records from the logged-in 企业微信 (WeCom) web page, thereby ingesting arbitrary user-generated table content from a third-party web app which the agent will read and act on.