SurfSense/surfsense_backend/app/agents/new_chat
CREDO23 db8bffab38 perf(prompt-cache): enable Azure prompt_cache_key routing hint
Splits the OpenAI-family gate into per-param predicates so AZURE and
AZURE_OPENAI configs now receive prompt_cache_key for backend routing
affinity (Microsoft auto-caches GPT-4o+ deployments at >=1024 tokens;
the key clusters same-prefix requests on the same GPU pool and raises
hit rate on turn 2+). prompt_cache_retention stays opted out for Azure
because litellm 1.83.14's Azure transformer would drop it silently;
revisit when Azure's supported params list is updated.
2026-05-20 11:58:15 +02:00
..
middleware perf(kb-planner): route internal planner calls to dedicated small/fast LLM 2026-05-20 11:42:52 +02:00
plugins chore: cleaned comments slop 2026-04-28 23:52:37 -07:00
prompts Wire Linear and Slack specialists and prompt routing. 2026-04-29 20:51:06 +02:00
skills feat: updated agent harness 2026-04-28 09:22:19 -07:00
subagents Merge upstream/dev into feature/multi-agent 2026-05-05 01:44:46 +02:00
tools perf(mcp): add per-call, discovery, and oauth-refresh timing logs 2026-05-19 21:29:56 +02:00
__init__.py cloud: added openrouter integration with global configs 2026-04-15 23:46:29 -07:00
agent_cache.py feat: implement agent caches and fix invalid prompt cache configs 2026-05-03 06:03:40 -07:00
chat_deepagent.py perf(kb-planner): route internal planner calls to dedicated small/fast LLM 2026-05-20 11:42:52 +02:00
checkpointer.py feat: implement connection pooling for AsyncPostgresSaver in checkpointer 2026-02-05 17:32:43 -08:00
context.py feat: improved document, folder mentions rendering 2026-05-09 22:15:51 -07:00
document_xml.py feat: updated file management for main agent 2026-04-28 04:32:52 -07:00
errors.py chore: cleaned comments slop 2026-04-28 23:52:37 -07:00
feature_flags.py refactor(chat): stream agent events via stream_output and remove parity v2 flag 2026-05-07 19:40:10 +02:00
filesystem_backends.py feat: updated file management for main agent 2026-04-28 04:32:52 -07:00
filesystem_selection.py feat(filesystem): refactor local filesystem handling to use mounts instead of root paths, enhancing mount management and path normalization 2026-04-24 05:59:21 +05:30
filesystem_state.py subagents/knowledge_base: wire KB specialist into orchestrator (renderer/projector split, FS middleware stack, cloud-mode gating) 2026-05-11 20:43:44 +02:00
llm_config.py feat: fixed vision/image provider specific errors and fixed podcast/video streaming 2026-05-02 19:18:53 -07:00
memory_extraction.py refactor(memory): streamline memory extraction by utilizing extract_text_content utility 2026-05-02 16:10:30 +05:30
mention_resolver.py feat: improved document, folder mentions rendering 2026-05-09 22:15:51 -07:00
path_resolver.py feat: updated file management for main agent 2026-04-28 04:32:52 -07:00
permissions.py chore: cleaned comments slop 2026-04-28 23:52:37 -07:00
plugin_loader.py chore: cleaned comments slop 2026-04-28 23:52:37 -07:00
prompt_caching.py perf(prompt-cache): enable Azure prompt_cache_key routing hint 2026-05-20 11:58:15 +02:00
sandbox.py feat: added ai file sorting 2026-04-14 01:43:30 -07:00
state_reducers.py feat: improved agent streaming 2026-04-29 07:20:31 -07:00
system_prompt.py Wire Linear and Slack specialists and prompt routing. 2026-04-29 20:51:06 +02:00
utils.py inject tools at runtime 2025-12-20 18:35:39 +02:00