/

CVE-2026-90651

Severity:

HIGH RISK

(8.1)

Improper Certificate Validation (CWE-295) in Socket Firewall

CVE-2026-90651

Deferred

Vulnerability type

Category:

Improper Certificate Validation

CWE:

CWE-295

Impact:

Man-in-the-Middle Response Tampering

Severity:

HIGH RISK

(8.1)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

Socket Firewall may disable TLS certificate verification by default when its registry-mode configuration omits the relevant verification settings. An attacker who can intercept traffic between the firewall and the Socket API or an upstream package registry could present an untrusted certificate and modify responses, including package content or policy decisions.

Affected context

The vulnerability occurs in registry-mode deployments of Socket Firewall when the generated configuration leaves API and upstream TLS verification disabled. It is particularly risky when traffic crosses networks where an attacker could intercept or redirect connections, and when the firewall relies on upstream responses to make package allow or block decisions.

Exploit preconditions

  • Attack surface: Outbound HTTPS connections from Socket Firewall to the Socket API or an upstream package registry
  • Required access: Network positioning sufficient to intercept, redirect, or otherwise influence the relevant traffic
  • Trigger point: TLS connections initiated by the OpenResty/Lua HTTP client
  • Why exploitation is possible: The client accepts self-signed or otherwise untrusted certificates when certificate verification is disabled, allowing intercepted responses to be treated as authentic

Why this happens

The underlying issue is an insecure TLS configuration default. When certificate-verification settings are omitted, generated configuration values disable validation rather than requiring verification or failing closed.

This maps to CWE-295: Improper Certificate Validation, because the client does not verify that a peer certificate is trusted and correctly associated with the intended service. As a result, encrypted transport does not reliably provide endpoint authentication or protect response integrity.

Potential impact

  • An attacker may read or alter data exchanged with the Socket API or package registry.
  • Malicious package content could be substituted during upstream retrieval.
  • Allow or block decisions enforced by the firewall could be modified.
  • Downstream systems may consume tampered metadata or package artifacts.
  • The integrity of software supply-chain controls could be weakened.

General mitigation guidance

Mitigation generally focuses on making TLS verification mandatory and ensuring the client has access to an appropriate trusted certificate authority bundle.

Common defensive patterns include:

  • Enable certificate verification for both API and upstream connections by default.
  • Configure an explicit, managed trust store containing only appropriate certificate authorities.
  • Ensure generated client configuration emits the trusted-certificate setting required by the TLS library.
  • Fail closed when verification settings or trust material are missing or invalid.
  • Protect configuration files from unauthorized modification.
  • Prefer private connectivity, authenticated proxies, or mutually authenticated TLS where suitable.
  • Run registry integrations with only the network and filesystem privileges they require.

How teams detect issues like this

Organizations can identify this class of issue through:

  • Static analysis and configuration checks for disabled TLS verification, such as false-valued verify settings or permissive certificate callbacks.
  • Code review searches for HTTP clients that omit trust-store configuration or accept all certificates.
  • Tests that connect to endpoints using expired, self-signed, mismatched, or untrusted certificates and verify that connections fail.
  • Policy enforcement requiring certificate verification and approved trust stores in generated deployment configuration.
  • Network monitoring for unexpected registry endpoints, certificate changes, or intercepted outbound connections.
  • Automated security review tools that flag insecure TLS defaults and missing certificate-validation paths.

Sources

SourceLink
docs.socket.devdocs.socket.dev/docs/registry-mode-configuration-reference
github.comgithub.com/SocketDev/socket-registry-firewall/releases/tag/v2.0.0

Frequently asked

questions

It may be exploitable remotely when an attacker can occupy a network position that permits interception or redirection of traffic between Socket Firewall and its configured services. It is not generally sufficient to have access to the public application alone.

The attacker typically needs network-level influence over the relevant outbound connection, such as control of a compromised network segment, proxy, gateway, or routing path. They also need the firewall to communicate with certificate verification disabled.

Require TLS certificate verification for every outbound service connection, configure a validated trust store, and make missing or invalid trust material cause the connection to fail rather than fall back to acceptance.

Use secure-by-default client settings, enforce TLS policies in generated configuration and CI checks, test failure behavior with invalid certificates, and monitor outbound destinations and certificate changes.

Find out what's already
exploitable in your codebase.