CVE-2026-45373: CodeWhale SSRF Vulnerability via IPv6 Bracket Notation

Technical summary

CodeWhale implements SSRF protections that validate against hostnames resolving to private IPv6 addresses. However, the validation logic does not properly handle IPv6 addresses when they are formatted using the standard URL bracket notation (http://[::1]). This bypasses the hostname resolution check because the bracket-enclosed IPv6 literal is parsed differently than the expected validation logic accounts for. By sending requests to http://[::1] or similar private IPv6 addresses in bracket notation, an attacker circumvents the SSRF defenses and gains the ability to make arbitrary network requests from the CodeWhale process context. The vulnerability is classified under CWE-918 (Server-Side Request Forgery), indicating a failure in input validation or network access control. The vector indicates the attack requires network access, low complexity, no privileges, and user interaction, with impact scope change allowing potential access to confidential data.

Business impact

If CodeWhale is deployed in a cloud or containerized environment, or on a shared host, an attacker can exploit this to access internal APIs, cloud metadata services (AWS IMDSv1, Google Cloud metadata), or other sensitive internal services. This could lead to credential theft, configuration exposure, or lateral movement within the infrastructure. For development teams using CodeWhale as part of their CI/CD pipeline or local development workflow, this creates a supply-chain or insider risk if the tool is compromised or if malicious prompts are injected. The risk is elevated by the fact that CodeWhale runs in terminal environments where it may have broader network permissions than typical web applications.

Affected systems

CodeWhale versions prior to 0.8.26 are vulnerable. The vulnerability applies to any deployment of the tool on systems where the CodeWhale process has network access to internal services, cloud metadata endpoints, or other restricted resources. End users and development teams running CodeWhale locally or in CI/CD environments should prioritize assessment and patching. Since no vendor product listing was provided in the source data, affected deployments are those explicitly running the CodeWhale terminal agent.

Exploitability

Exploiting this vulnerability requires user interaction (the CVSS vector indicates UI:R), meaning an attacker would need to trick a CodeWhale user into making a request (e.g., through a crafted prompt or malicious configuration) that triggers the tool to fetch a URL with a bracketed IPv6 address. The attack does not require authentication or special privileges. The network-based attack vector and low complexity make it feasible for remote attackers if they can influence what URLs CodeWhale attempts to fetch. However, the requirement for user interaction provides some friction—exploitation is not fully automatic without social engineering or prompt injection.

Remediation

Upgrade CodeWhale to version 0.8.26 or later, which includes a fix for the IPv6 bracket notation SSRF bypass. Organizations should inventory all deployments of CodeWhale, particularly in CI/CD pipelines and development environments. After patching, validate that SSRF protections are functioning by testing with IPv6 localhost notation. If you cannot upgrade immediately, implement network segmentation to restrict the CodeWhale process's access to internal services and metadata endpoints using firewall rules or container network policies.

Patch guidance

Apply the upgrade to CodeWhale 0.8.26 or later. Verify the patch by checking the tool's version output (typically via a --version flag or similar). Since CodeWhale is a terminal agent, patches should be deployed to all machines, CI/CD runners, and containers where the tool is installed. Ensure any local caches or pinned versions in dependency management files (if applicable) are updated. Test in a non-production environment first to confirm the upgrade does not break existing workflows.

Detection guidance

Monitor for network requests from CodeWhale processes to private IPv6 addresses or localhost (::1, ::ffff:127.0.0.1) using URLs with bracket notation. Review proxy logs and host-level network monitoring for outbound requests from CodeWhale to unexpected internal services or cloud metadata endpoints. If you use endpoint detection and response (EDR) tools, create rules to flag network activity initiated by the CodeWhale process to restricted ranges. Inspect CodeWhale's command history or prompt logs if available to detect suspicious or unusual requests.

Why prioritize this

This vulnerability should be prioritized due to its HIGH severity (CVSS 7.4), potential for confidential data exposure, and the broad attack surface in modern development environments where CodeWhale may be deployed. The SSRF vulnerability class is well-understood and commonly exploited in supply-chain and infrastructure attacks. The requirement for user interaction provides moderate mitigation, but the ease of crafting a malicious prompt or configuration makes this a realistic threat. Immediate patching in production and sensitive development environments is recommended.

Risk score, explained

The CVSS 7.4 score reflects a HIGH severity vulnerability due to several factors: (1) Network-based attack vector with no authentication requirement, (2) Low attack complexity, (3) Potential for confidential information disclosure (C:H), and (4) Scope change, indicating the impact could extend beyond the vulnerable component to other systems accessible by CodeWhale. However, the score is not CRITICAL because exploitation requires user interaction and does not enable code execution or system availability impact directly on CodeWhale itself.

Frequently asked questions