yucheng 35a1d017dc feat(guardrails): send MCP tool metadata to Agent 365 and treat unevaluated Defender verdicts as unavailable ... Carry the listed tool's description and inputSchema from MCPServerManager through the pre-call and during-call hook request objects into the Agent 365 evaluate payload, omitting them when the tool was never listed. An allowed verdict whose defender.status is not Evaluated (Skipped, FailedOpen, missing) now follows the unreachable_fallback policy instead of counting as a scanned allow. The conversationId prefers the proxy-owned litellm_call_id over caller-controlled mcp-session-id headers. Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>		2026-09-11 02:08:01 +00:00
..
helpers	Merge remote-tracking branch 'origin/litellm_internal_staging' into litellm_openai_error_payload_non_llm_routes	2026-09-08 11:09:54 -07:00
prisma_and_spend	Merge pull request #39556 from BerriAI/litellm_fix_spend_log_flush_event_loop_binding	2026-09-08 15:59:11 -07:00
proxy_logging	feat(guardrails): send MCP tool metadata to Agent 365 and treat unevaluated Defender verdicts as unavailable	2026-09-11 02:08:01 +00:00
__init__.py	test(proxy/utils): pin bottom-of-file helper behavior (#29509)	2026-06-02 17:45:19 -07:00