i/Aether
1
0
Fork 0
mirror of https://github.com/fawney19/Aether.git synced 2026-09-12 14:10:19 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix(ws): harden Responses connection lifecycle

Browse Source 
Revalidate control policy per turn, isolate downstream credentials, and make planner/turn ownership cancellation-safe.

Preserve opaque protocol events, align configurable timeout semantics, and extend end-to-end security and settlement coverage.
This commit is contained in:
ZheFox
2026-08-17 18:50:29 +08:00
parent 4a0775c4ea
commit c8118edf36
42 changed files with 4017 additions and 1276 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/WebSocket-Mode.md
+1 -1
View File
@@ -8,7 +8,7 @@ WebSocket mode is compatible with both Zero Data Retention (ZDR) and `store=fals
WebSocket mode is most useful when a workflow involves many model-tool round trips (for example, agentic coding or orchestration loops with repeated tool calls).
Because the connection stays open and each turn sends only incremental input, WebSocket mode reduces per-turn continuation overhead and improves end-to-end latency across long chains. For rollouts with 20+ tool calls, we have seen up to roughly 40% faster end-to-end execution.
Because the connection stays open and each turn sends only incremental input, WebSocket mode reduces per-turn continuation overhead and improves end-to-end latency across long chains. The [OpenAI WebSocket-mode guide](https://developers.openai.com/api/docs/guides/websocket-mode) reports up to roughly 40% faster end-to-end execution for workloads with 20 or more tool calls; this is an upstream product claim, not an Aether benchmark.
## Connect and create responses