5.14.0 至 5.15.2(不含 5.15.2)版本的 Network-AI 在 ClaudeHookBridge 的 PreToolUse 钩子中存在安全控制绕过漏洞。该漏洞仅针对第一个存在的候选字段匹配拒绝模式(denyPatterns)。攻击者可通过提示词注入(prompt injection)的代理程序,在 benign(合法)的 file_path 或 command 字段旁边,于诸如“Write content”或 MCP 脚本参数等字段中植入危险载荷,从而绕过拒绝模式检查,写入 Web shell
| Vendor | Product | Version Range | Status |
|---|---|---|---|
| Jovancoding | Network-AI | 5.14.0< 5.15.2 |
affected |
5.15.2 |
unaffected |
Although we use advanced large model technology, its output may still contain inaccurate or outdated information.Shenlong tries to ensure data accuracy, but please verify and judge based on the actual situation.
| Vendor | Product | Affected Versions | CPE | Subscribe |
|---|---|---|---|---|
| Jovancoding | Network-AI | 5.14.0 ~ 5.15.2 | - |
|
| # | POC Description | Source Link | Shenlong Link |
|---|
No public POC found.
Login to generate AI POC| CVE-2026-109330 | 4.3 MEDIUM | Network-AI 5.0.0 before 5.15.2 Cross-Site WebSocket Hijacking via DashboardServer |
| CVE-2026-109332 | 3.1 LOW | Network-AI before 5.15.2 CORS Origin Validation Error in MCP SSE Transport |
No comments yet