Rodrigo Rodriguez (Pragmatismo) f9178e947e Use max_completion_tokens for GPT-5 models		2026-04-16 16:21:16 -03:00
..
context
llm_models	Add AzureGPT5 client and provider detection	2026-04-16 15:07:23 -03:00
prompt_manager
bedrock.rs	fix: normalize roles to system for bedrock and vertex LLM providers	2026-04-14 13:44:12 -03:00
cache.rs	fix: whatsapp dynamic routing and openai tool call accumulation	2026-03-10 17:19:17 -03:00
cache_test.rs
claude.rs	fix: remove hardcoded salesianos, strip think tags globally, block reasoning_content leak	2026-04-13 09:04:22 -03:00
episodic_memory.rs	fix: update Minimax and DeepSeek handlers to strip unclosed thinking tags	2026-04-14 16:13:41 -03:00
glm.rs	fix: normalize episodic/compact roles to system in all LLM providers	2026-04-14 13:47:18 -03:00
hallucination_detector.rs
kimi.rs	Add AzureGPT5 client and provider detection	2026-04-16 15:07:23 -03:00
llm.md	Fix tenant-org-bot relationship and CRM lead form	2026-03-12 18:19:18 -03:00
local.rs	fix: extract base URL for embedding health checks	2026-04-12 19:33:35 -03:00
mod.rs	Use max_completion_tokens for GPT-5 models	2026-04-16 16:21:16 -03:00
observability.rs	Read Drive config from Vault at runtime with fallback defaults	2026-03-17 00:00:36 -03:00
rate_limiter.rs
smart_router.rs	Fix DETECT: use bot-specific DB pool, add anonymous auth when directory disabled	2026-04-06 13:37:23 -03:00
vertex.rs	fix: normalize roles to system for bedrock and vertex LLM providers	2026-04-14 13:44:12 -03:00