by Kastro
Paper/Archive/CVE-2026-75062

CVE-2026-75062

Google langfun — Eval Injection in google/langfun via default lf.query protocol

EPSS 0.4%public exploitCVSS 9.2
critical.is priorityof 100
43
CVSS 9.2 (CNA)9.2 / 10 × 3027.6Not in CISA KEV+300.0EPSS percentile 28.7.287 × 205.7Exploit: SSVC poc+1010.0SSVC exploitation: poc+100.0
Description · from the CVE record

Improper Neutralization of Directives in Dynamically Evaluated Code ('Eval Injection') in the default lf.query Python protocol in Google langfun versions prior to 0.1.2 allows remote unauthenticated attackers to execute arbitrary Python code in the context of the host application via crafted prompt inputs that cause the model to generate executable Python expressions evaluated without a sandbox.

Affected

ProductAffectedFixed inSource
langfun0.0.1 – < 0.1.20.1.2CNA

Scores by source

Google (CNA)9.2CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N
CISA ADPSSVCExploitation: poc · Automatable: no · Technical impact: total

References

refgithub.com/google/langfunexploitgithub.com/google/langfun/issues/725