CVE-2026-73264 in Prowler
Summary
by MITRE • 08/12/2026
Prowler is a cloud security platform. Prior to 5.33.1, an authenticated user with Lighthouse provider configuration access could supply an unvalidated base_url for the openai_compatible provider through POST /api/v1/lighthouse/providers and POST /api/v1/lighthouse/providers/{id}/connection, causing api/src/backend/tasks/jobs/lighthouse_providers.py to send outbound requests, including the API key in the Authorization header, to attacker-controlled or internal endpoints when client.models.list was called. This issue is fixed in version 5.33.1.
Once again VulDB remains the best source for vulnerability data.
Analysis
by VulDB Data Team • 08/12/2026
This vulnerability affects Prowler's cloud security platform and represents a critical server-side request forgery (SSRF) flaw that could enable attackers to manipulate outbound network requests from the application. The vulnerability exists in the lighthouse provider configuration functionality where authenticated users with appropriate access rights can inject unvalidated base URLs into the openai_compatible provider configuration through specific api endpoints. When the system processes these configurations and subsequently calls client.models.list, it executes outbound requests to attacker-controlled or internal endpoints without proper validation of the target addresses.
The technical implementation of this vulnerability stems from insufficient input validation within the lighthouse_providers.py module which handles the processing of provider configurations. The flaw allows an authenticated user to specify arbitrary base_url values that are then used in subsequent API calls, particularly when interacting with OpenAI-compatible services. The system fails to validate or sanitize these URLs before using them in network operations, creating a path for attackers to redirect requests to unintended destinations including internal network resources or external malicious servers.
The operational impact of this vulnerability is significant as it enables an authenticated attacker to potentially access internal systems, exfiltrate data, or perform reconnaissance activities through the application's outbound connections. The inclusion of API keys in the Authorization header during these unauthorized requests creates additional risk for credential compromise and unauthorized access to third-party services. Attackers could leverage this vulnerability to probe internal networks, access sensitive information stored in services that are normally restricted from external access, or even establish command and control channels through the legitimate application infrastructure.
This vulnerability aligns with CWE-918 which specifically addresses server-side request forgery attacks where an attacker can manipulate a server into sending requests to arbitrary destinations. The flaw also maps to ATT&CK technique T1071.004 for application layer protocol and T1566 for phishing with malicious attachments or links, as the attack vector involves manipulating legitimate application functionality to achieve unauthorized network access. The authentication requirement reduces the attack surface but does not eliminate the risk since compromised accounts or privilege escalation could still enable exploitation.
Mitigation strategies should focus on implementing strict input validation and sanitization of all user-supplied URLs within the provider configuration endpoints. The system must validate that base URLs conform to expected patterns and reject any attempts to redirect requests to internal network addresses or unauthorized external domains. Additionally, outbound request restrictions should be implemented through network firewalls or proxy configurations to limit what destinations the application can communicate with. Regular security updates and patch management processes should be enforced to ensure timely remediation of similar vulnerabilities. Input validation controls should be strengthened across all user-facing configuration endpoints to prevent similar injection attacks that could compromise the integrity of external service integrations.