FlowiseAI/Flowise 存在服务器端请求伪造 (SSRF) 漏洞

攻击者可以利用此SSRF漏洞将Flowise服务器作为代理，绕过防火墙和安全策略，访问内部网络中的敏感资源。攻击者可以扫描内部网络，发现未公开的服务和应用程序，并尝试利用这些服务的漏洞。此外，攻击者还可以利用此漏洞访问内部管理界面，执行未经授权的操作，例如修改配置、获取敏感数据或控制系统。由于Flowise通常部署在与内部网络隔离的环境中，因此此漏洞可能导致内部网络的暴露，增加攻击面。

Organizations deploying Flowise in environments with internal web services or APIs are at risk. Shared hosting environments where Flowise instances share the same network infrastructure are particularly vulnerable, as an attacker could potentially pivot from a compromised Flowise instance to other internal services. Legacy Flowise configurations that haven't been regularly updated are also at increased risk.

• nodejs: Monitor process execution for unusual outbound network connections originating from the Flowise process. Use ps aux | grep flowise to identify the process and then netstat -anp | grep <flowise_pid> to check connections. • generic web: Examine access logs for requests to /api/v1/fetch-links with unusual or internal IP addresses in the URL. Use grep '/api/v1/fetch-links' access.log | grep <internal_ip>. • generic web: Check response headers for signs of internal resource exposure. Look for headers that reveal internal server information or redirect to internal services.

1. 2025-09-15Disclosure

   disclosure

1. 已保留2025-09-17
2. 发布日期2025-09-15
3. 修改日期2025-09-22
4. EPSS 更新日期2026-03-23

为了缓解CVE-2025-59527的影响，建议立即升级Flowise至3.0.6版本或更高版本。如果无法立即升级，可以考虑以下临时缓解措施：限制/api/v1/fetch-links端点的访问权限，只允许来自受信任来源的请求。实施严格的网络策略，阻止Flowise服务器访问不必要的内部网络资源。使用Web应用程序防火墙（WAF）来检测和阻止SSRF攻击。监控Flowise服务器的日志，查找可疑活动。