mlc-ai xgrammar

• CVE-2026-25048 High 7.5 xgrammar: Multi-layer nesting causes DoS 1.3%
• CVE-2025-58446 Medium 6.9 xgrammar vulnerable to denial of service by huge enum grammar 0.6%
• CVE-2025-57809 High 7.7 XGrammar affected by Denial of Service by infinite recursion grammars 0.5%
• CVE-2025-32381 Medium 6.5 Denial of Service by abusing xgrammar unbounded cache in memory 0.5%