Ayuriel/openclaw
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Microsoft Foundry: use active model hint at runtime

Browse Source 
Prefer the currently selected model hint during runtime auth refresh so switching Foundry deployments cannot reuse stale onboarding metadata and route requests to the wrong GPT-5 or non-GPT-5 endpoint.
This commit is contained in:
haxudev 2026-03-19 16:52:32 +08:00
parent 27fe74646c
commit 0c1efec1c9
1 changed files with 6 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
extensions/microsoft-foundry/index.ts
View File

@ -1079,7 +1079,12 @@ export default definePluginEntry({
: typeof metadata?.modelId === "string" && metadata.modelId.trim().length > 0
? metadata.modelId.trim()
: ctx.modelId;
const modelNameHint = resolveConfiguredModelNameHint(modelId, metadata?.modelName ?? ctx.model.name);
const activeModelNameHint =
ctx.modelId === metadata?.modelId ? metadata?.modelName : undefined;
const modelNameHint = resolveConfiguredModelNameHint(
modelId,
ctx.model.name ?? activeModelNameHint,
);
const endpoint =
typeof metadata?.endpoint === "string" && metadata.endpoint.trim().length > 0
? metadata.endpoint.trim()