PatchSiren

PatchSiren cyber security CVE debrief

CVE-2026-107685 agentjido CVE debrief

CVE-2026-107685 Improper Certificate Validation vulnerability in req_llm allows on-path attackers to read and modify OpenAI WebSocket sessions. The vulnerability affects req_llm versions from 1.9.0 before 1.29.0. Defenders should assess exposure and prioritize verification and remediation. The vulnerability is caused by the WebSocket streaming client in agentjido req_llm not trusting TLS certificates, allowing an on-path attacker to read and modify OpenAI WebSocket sessions, including the authorization bearer token.

Vendor
agentjido
Product
req_llm
CVSS
CRITICAL 9.1
CISA KEV
Not listed in stored evidence
Original CVE published
2026-10-23
Original CVE updated
2026-11-02
Advisory published
2026-10-23
Advisory updated
2026-11-02

Who should care

Defenders responsible for req_llm deployments, OpenAI WebSocket sessions, and TLS certificate configurations should assess exposure and prioritize verification and remediation. They should review and update TLS certificate configurations for req_llm, and ensure proper certificate validation for OpenAI WebSocket sessions.

Why it matters

CVE-2026-107685 is a critical vulnerability in req_llm that allows on-path attackers to read and modify OpenAI WebSocket sessions. Defenders should prioritize verifying exposure, updating vulnerable versions, and ensuring proper certificate validation to prevent exploitation.

  • On-path attackers can read and modify OpenAI WebSocket sessions
  • Authorization bearer tokens can be compromised
  • req_llm versions between 1.9.0 and 1.29.0 are vulnerable
  • Proper certificate validation is required to prevent exploitation

Technical summary

The WebSocket streaming client in agentjido req_llm has an Improper Certificate Validation vulnerability, allowing an on-path attacker to read and modify OpenAI WebSocket sessions, including the authorization bearer token, via a TLS certificate the client does not trust. The vulnerability affects req_llm versions from 1.9.0 before 1.29.0. The client starts WebSockex without TLS options, and WebSockex defaults to insecure: true, which connects with verify: :verify_none. The option allowlist also drops any caller-supplied cacerts or ssl_options, so an application cannot turn verification on.

Defensive priority

Defenders should prioritize verifying exposure of req_llm versions between 1.9.0 and 1.29.0, and assess the use of OpenAI WebSocket sessions.

Recommended defensive actions

  • Verify req_llm versions between 1.9.0 and 1.29.0 are updated to 1.29.0 or later
  • Assess the use of OpenAI WebSocket sessions and ensure proper certificate validation
  • Review and update TLS certificate configurations for req_llm
  • Confirm whether affected product deployments exist in managed environments and assign an owner for follow-up
  • Review the supplied official advisory or CVE record to validate affected scope, severity, and vendor guidance
  • Plan vendor-supported updates or mitigations through normal change control where exposure is confirmed
  • Check relevant monitoring, detection, and logs for exposed assets that need extra review

Evidence notes

The CVE record and NVD entry provide details on the vulnerability, including its description, affected versions, and references. The vulnerability has a CVSS score of 9.1 and is considered critical. The OpenAI Responses, Realtime, and Codex WebSocket paths reach this client and send the API key or OAuth access token in the authorization header. HTTP streaming through Finch and the HTTP/2 duplex session through Mint verify certificates and are not affected.

Sources and references

Verified primary and authoritative sources

  • CVE-2026-107685 CVE Program record

    Publisher, destination, and source semantics verified

    URL: https://www.cve.org/CVERecord?id=CVE-2026-107685

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

  • CVE-2026-107685 NVD vulnerability detail

    Publisher, destination, and source semantics verified

    URL: https://nvd.nist.gov/vuln/detail/CVE-2026-107685

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

Supplemental references

  • Source reference

    Unverified legacy reference

    URL: https://cna.erlef.org/cves/CVE-2026-107685.html

    6b3ad84c-e1a6-4bf7-a703-f496b71e49db

  • Source reference

    Unverified legacy reference

    URL: https://github.com/agentjido/req_llm/commit/a389d718640ceb2d541a8f540ab954cbf5ef8313

    6b3ad84c-e1a6-4bf7-a703-f496b71e49db

  • Source reference

    Unverified legacy reference

    URL: https://github.com/agentjido/req_llm/commit/c7264fc0217fbacfb5562234ff790b8e4462de76

    6b3ad84c-e1a6-4bf7-a703-f496b71e49db

  • Source reference

    Unverified legacy reference

    URL: https://github.com/agentjido/req_llm/security/advisories/GHSA-6vww-5m28-f26v

    6b3ad84c-e1a6-4bf7-a703-f496b71e49db

  • Source reference

    Unverified legacy reference

    URL: https://osv.dev/vulnerability/EEF-CVE-2026-107685

    6b3ad84c-e1a6-4bf7-a703-f496b71e49db

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.