fix: OWASP AI Agent Security hardening for chat and MCP endpoints#1073

Merged

BenjaminMichaelis merged 12 commits intomainfrom

agents/ai-agent-security-review

May 9, 2026

Commits on May 7, 2026

fix: implement OWASP AI Agent security findings (#1064-#1071)
BenjaminMichaelis
committed
fix: address multi-agent review findings (security hardening iteration 2)
BenjaminMichaelis
committed
fix: cleanup from second agent review pass
BenjaminMichaelis
committed

Commits on May 8, 2026

fix: address PR #1073 review feedback (security hardening iteration 3)
BenjaminMichaelis
committed
fix: narrow token-limit exception type and clean up minor review findings
BenjaminMichaelis
committed
fix: per-call MemoryCacheEntryOptions and add TokenLimitExceededException docs
BenjaminMichaelis
committed
test: DRY up McpApiTokenServiceTests and ResponseIdValidationServiceTests
BenjaminMichaelis
committed
fix: address PR #1073 second-round feedback (security hardening iteration 4)
BenjaminMichaelis
committed
fix: resolve streaming responseId duplication and ownership gaps
BenjaminMichaelis
committed
fix: ResponseIdValidationService owns its IMemoryCache; remove shared SizeLimit; remove dead cleanupOldSessions JS
BenjaminMichaelis
committed

Commits on May 9, 2026