gzh-prohibited-word（公众号违禁词检测）

• Analyzed: May 10, 2026, 6:49 AM
• Findings: 1
• Version: 1.0.0
• Hash: e8961ea816e2dd16541aa2016c741af6ddefa94ce96aebb6a8afdb623971a61b
• Source: palm
• External report: View on VirusTotal

Findings (1)

The skill performs content moderation by sending user-provided text, files, or URL content to an external API. While the logic aligns with its stated purpose, the script `scripts/check_sensitive_words.py` employs high-risk networking techniques, including a manual socket/SSL implementation that intentionally omits Server Name Indication (SNI) to bypass network filtering and explicitly disables SSL certificate verification (`ssl.CERT_NONE`), making the connection vulnerable to Man-in-the-Middle attacks. Furthermore, the instructions in `references/core_workflow.md` mandate that the agent send extracted content to a configurable external endpoint and write files to the local disk, which could lead to data exposure if sensitive files are processed. These behaviors represent significant security vulnerabilities and bypass mechanisms rather than clear evidence of intentional malice.