CVE-2026-111391 sgl-project CVE debrief

Who should care

Defenders responsible for SGLang deployments should assess exposure and prioritize remediation to prevent code injection attacks. This includes verifying exposure by checking if the /dumper/configure endpoint is accessible and if the DUMPER_SERVER_PORT is set. Additionally, defenders should monitor for suspicious activity on the /dumper/configure endpoint and implement compensating controls to prevent code injection attacks.

Why it matters

CVE-2026-111391 is a critical vulnerability in SGLang versions 0.5.10 through 0.5.22, allowing remote code injection via the /dumper/configure endpoint. Defenders should prioritize verifying exposure and applying remediation to prevent code injection attacks.

• Verify exposure and apply remediation to prevent code injection attacks.
• Monitor for suspicious activity on the /dumper/configure endpoint.
• Implement compensating controls to prevent code injection attacks.
• Assess the impact of a potential code injection attack on the scheduler process.

Technical summary

The vulnerability is caused by a code injection issue in the _evaluate_filter() function, which allows attackers to execute Python code by escaping the eval() sandbox. The issue is triggered when sending a crafted filter expression to the unauthenticated /dumper/configure endpoint when DUMPER_SERVER_PORT is set. This allows remote attackers to execute Python code in the scheduler process on the next forward pass. The vulnerability has a CVSS score of 9.2 and is considered critical. Defenders should prioritize verifying exposure and applying remediation to prevent code injection attacks.

Defensive priority

Defenders should prioritize verifying exposure and applying remediation to prevent code injection attacks.

Recommended defensive actions

• Verify exposure by checking if the /dumper/configure endpoint is accessible and if the DUMPER_SERVER_PORT is set.
• Apply remediation by updating SGLang to a version that fixes the code injection vulnerability.
• Monitor for suspicious activity on the /dumper/configure endpoint.
• Implement compensating controls to prevent code injection attacks.
• Assess the impact of a potential code injection attack on the scheduler process.
• Review relevant monitoring, detection, and logs for exposed assets that need extra review.
• Track exceptions, retest remediated assets, and close the item only after evidence is documented.

Evidence notes

The vulnerability is caused by a code injection issue in the _evaluate_filter() function, which allows attackers to execute Python code by escaping the eval() sandbox. The issue is triggered when sending a crafted filter expression to the unauthenticated /dumper/configure endpoint.

Sources and references

Methodology and review provenance

AI-assisted synthesis based on stored public vulnerability evidence. System validation, approval state, and publication status do not by themselves establish human review of this revision. PatchSiren helps prioritize defensive review and does not prove exposure or remediation on any system.