CVE-2026-105148 SciPhi-AI CVE debrief

Who should care

Defenders responsible for SciPhi-AI R2R deployments, especially those using versions up to 3.6.6, should assess the potential impact of this vulnerability and prioritize verification and mitigation efforts.

Why it matters

CVE-2026-105148 is a server-side request forgery vulnerability in SciPhi-AI R2R up to 3.6.6 that can be exploited remotely. Defenders should verify the presence of this vulnerability in their deployments and assess the potential impact.

• Verification of affected versions and deployments is required to determine the scope of potential impact.
• Server-side request forgery could lead to unauthorized requests and potential data breaches.
• Defenders should assess the potential impact on their specific deployment contexts.

Technical summary

The vulnerability affects the Retrieval Completion API Endpoint in SciPhi-AI R2R up to 3.6.6, allowing for server-side request forgery through manipulation of the generation_config.api_base argument in the file py/shared/abstractions/llm.py. This could lead to unauthorized requests and potential data breaches. Defenders should prioritize verifying the presence of this vulnerability in their SciPhi-AI R2R deployments, especially those using versions up to 3.6.6, and assess the potential impact of server-side request forgery.

Defensive priority

Defenders should prioritize verifying the presence of this vulnerability in their SciPhi-AI R2R deployments, especially those using versions up to 3.6.6, and assess the potential impact of server-side request forgery.

Recommended defensive actions

• Verify the presence of this vulnerability in SciPhi-AI R2R deployments, especially those using versions up to 3.6.6.
• Assess the potential impact of server-side request forgery on the deployment.
• Consider applying patches or mitigations if available.
• Review compensating controls for exposed systems while remediation is scheduled and verified.
• Check 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 NVD entry provide information about the vulnerability, but the vendor did not respond to the disclosure. The exploit is publicly available, but there is no information on its usage.

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.