mmc/lingma-openai-gateway
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix: align agent payload semantics with VSCode tool flow

Browse Source 
Force OpenAI tooling-context requests into agent mode and align Lingma ask payload fields for agent requests so server-side tool path matches VSCode semantics.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
This commit is contained in:
GitHub Actions
2026-04-19 23:19:52 +08:00
parent e208025f35
commit c67a9c3d61
3 changed files with 9 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
app/main.py
View File

@@ -504,13 +504,13 @@ async def v1_chat_completions(req: ChatCompletionsRequest, request: Request):
# 1. Reuse the upstream sessionId so Lingma/Qwen hits its KV cache.
# 2. Send only the new user message instead of the whole history.
# 3. Stick the request to the pool instance that originally served it.
ask_mode = settings.default_ask_mode
if req.model.lower() in {"lingma-agent", "agent"}:
ask_mode = "agent"
tool_config = _openai_tool_config(req)
has_tooling_context = _openai_has_tooling_context(req, messages_dump)
ask_mode = settings.default_ask_mode
if req.model.lower() in {"lingma-agent", "agent"} or has_tooling_context:
ask_mode = "agent"
reuse_eligible = (
session_cache.enabled
and ask_mode == "chat"