PatchSiren

PatchSiren cyber security CVE debrief

CVE-2026-110807 sgl-project CVE debrief

A weakness has been identified in sgl-project SGLang up to 0.5.19. Affected by this vulnerability is an unknown functionality of the file python/sglang/srt/managers/io_struct.py of the component generate Endpoint. This manipulation of the argument session_params causes denial of service. The attack is possible to be carried out remotely. The exploit has been made available to the public and could be used for attacks. The project was informed of the problem early through an issue report but has not responded yet.

Vendor
sgl-project
Product
SGLang
CVSS
MEDIUM 5.3
CISA KEV
Not listed in stored evidence
Original CVE published
2026-10-14
Original CVE updated
2026-10-15
Advisory published
2026-10-14
Advisory updated
2026-10-15

Who should care

Defenders responsible for systems using SGLang, particularly those exposed to remote attacks, should assess exposure and prioritize verification and potential remediation. Defenders need to verify affected versions and assess exposure. The vulnerability allows for remote denial of service attacks, impacting system availability. Public exploits are available, increasing the urgency for remediation. Remediation priority is elevated due to the potential for

Why it matters

CVE-2026-110807 is a denial of service vulnerability in sgl-project SGLang up to 0.5.19, allowing remote attacks. Defenders should verify versions, assess exposure, and prioritize remediation due to public exploits.

  • Remote denial of service attacks are possible, impacting system availability.
  • Defenders need to verify affected versions and assess exposure.
  • Public exploits are available, increasing the urgency for remediation.
  • Remediation priority is elevated due to the potential for exploitation.

Technical summary

The vulnerability is located in the io_struct.py file of the SGLang generate Endpoint, specifically in the handling of the session_params argument. This allows for remote denial of service attacks. The vulnerability affects sgl-project SGLang up to 0.5.19. The attack is possible to be carried out remotely. The exploit has been made available to the public and could be used for attacks. The project was informed of the problem early through an issue report but has not responded yet. Defenders should verify versions, assess exposure, and prioritize remediation due to public exploits.

Defensive priority

Defenders should prioritize verifying the affected versions of SGLang and assessing exposure, as the vulnerability allows for remote denial of service attacks.

Recommended defensive actions

  • Verify the version of SGLang being used and check if it is within the affected range (up to 0.5.19).
  • Assess exposure to the vulnerability, particularly in remote attack scenarios.
  • Monitor for public exploits and adjust defensive priorities accordingly.
  • Consider compensating controls or workarounds until an official patch is available.
  • 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.
  • Confirm whether affected product deployments exist in managed environments and assign an owner for follow-up.

Evidence notes

The CVE record and source item provide details about the vulnerability, including its description, affected versions, and references to external sources. The vulnerability is located in the io_struct.py file of the SGLang generate Endpoint, specifically in the handling of the session_params argument. This allows for remote denial of service attacks. The project was informed of the problem early through an issue report but has not responded yet. The exploit has been made available to the public and could be used for attacks. Defenders

Sources and references

Verified primary and authoritative sources

  • CVE-2026-110807 CVE Program record

    Publisher, destination, and source semantics verified

    URL: https://www.cve.org/CVERecord?id=CVE-2026-110807

    CVE Program - Official CVE Program record with source-provided CVE metadata.

  • CVE-2026-110807 NVD vulnerability detail

    Publisher, destination, and source semantics verified

    URL: https://nvd.nist.gov/vuln/detail/CVE-2026-110807

    NIST National Vulnerability Database - Official NIST NVD detail page and source-specific vulnerability assessment.

Supplemental references

  • sgl-project SGLang generate Endpoint io_struct.py denial of service

    Unverified legacy reference

    URL: https://raw.githubusercontent.com/CVEProject/cvelistV5/main/cves/2026/110xxx/CVE-2026-110807.json

    cve_program_cvelist_v5

  • Source reference

    Unverified legacy reference

    URL: https://vuldb.com/vuln/420278

    Supplemental source - vdb-entry, technical-description

  • Source reference

    Unverified legacy reference

    URL: https://vuldb.com/vuln/420278/cti

    Supplemental source - signature, permissions-required

  • Source reference

    Unverified legacy reference

    URL: https://vuldb.com/cve/CVE-2026-110807

    Supplemental source - third-party-advisory

  • Source reference

    Unverified legacy reference

    URL: https://vuldb.com/submit/1004351

    Supplemental source - third-party-advisory

  • Source reference

    Unverified legacy reference

    URL: https://github.com/sgl-project/sglang/issues/41467

    Supplemental source - issue-tracking

  • Source reference

    Unverified legacy reference

    URL: https://github.com/sgl-project/sglang/pull/41535

    Supplemental source - issue-tracking, patch

  • Source reference

    Unverified legacy reference

    URL: https://gist.github.com/Yunzez/28820d220c52de4810fd5fb2d1dcf409

    Supplemental source - exploit

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.